Executive Summary
With the emergence of new multi discipline cycling sports such as cyclo-cross, the benefit of designing of new generation bicycles has rapidly increased. In these new cross bread versions of the sport of cycling the objective of maximising the strength to weight ratio is of the utmost importance, more often than not at the expense of price. In particular, the bicycle frame warrants a high level of attention due to its centralised position, relatively large size and exposure to a range of stresses and strains.
To achieve this goal bicycle designers are ever looking for improved materials to manufacturing the components from. This report outlines the results of a product development process used to identify the ideal material and manufacturing process used for constructing a high performance cyclo-cross bicycle. The report details the properties and microstructure of the selected titanium alloy and the steps and influence of key process variables on the seamless tube forming and tube buttering processes. The report outlines the process control and inspection and test plans proposed to ensure product quality is maintained. In addition the overall operational costs for manufacturing the frame are covered to determine an economic batch size and the viability of such a business.
2
Table of Contents Executive Summary ................................................................................................................. 2 1. Product Development..................................................................................................... 4 1.1. Engineering Design and Analysis ................................................................................... 4 1.2. Material Selection Process ................................................................................................ 7 1.3. Ti-3Al-2.5V ........................................................................................................................... 11 2. Manufacturing Process ............................................................................................... 14 2.1. Process Selection ............................................................................................................... 14 2.1.1. Primary Processing: Kroll Process .................................................................................. 14 2.1.2. Primary Shaping Options .................................................................................................... 15 2.2. Seamless Tube Manufacturing...................................................................................... 18 2.2.1. Roll Piercing ............................................................................................................................. 18 2.2.2. Vacuum Annealing ................................................................................................................. 19 2.2.3. Pilgering ..................................................................................................................................... 19 2.2.4. Final Heat Treatment............................................................................................................ 20 2.2.5. Tube Butting ............................................................................................................................. 20 2.3. Combined Frame ............................................................................................................... 22 3. Operating Costs .............................................................................................................. 23 3.1. Key Assumptions ............................................................................................................... 23 3.2. Material Costs ..................................................................................................................... 23 3.3. Tooling Costs ....................................................................................................................... 23 3.4. Capital Costs ........................................................................................................................ 24 3.5. Overhead Costs ................................................................................................................... 25 3.6. Unit Cost Analysis .............................................................................................................. 25 3.7. Major Cost Factors............................................................................................................. 26 4. Process and Quality Control ...................................................................................... 27 4.1. Process Variables .............................................................................................................. 27 4.1.1. Temperature ............................................................................................................................ 27 4.1.2. Force ............................................................................................................................................ 28 4.1.3. Time ............................................................................................................................................. 28 4.2. Process Control .................................................................................................................. 29 4.3. Quality Control ................................................................................................................... 29 4.3.1. Inspection Process ................................................................................................................. 30 4.3.2. Testing Process ....................................................................................................................... 30 5. Component Joining ....................................................................................................... 31 5.1. Tungsten Inert Gas (TIG) Welding .............................................................................. 31 References ................................................................................................................................ 33
1. Product Development To determine the requirements of a high performance cyclo-cross bicycle the methodical product development process was utilised. The initial stage of the process involves defining the design problem and is done via analysis of the target customer needs. Metrics are developed for each of these needs to quantify them. After thorough benchmarking against current competitor products a series of design targets are developed. These targets form the basis of the ranking matrix used to select the best material type for the frame.
1.1. Engineering Design and Analysis As show in Figure 1 a bicycle frame consists of a number of different tubes. The frame can be constructed as one solid piece or via joining the number of separate parts. In many designs each of the tube sections have a different geometry, in order to capitalise of mass savings in low stress areas and to add aerodynamic capabilities [1].
Figure 1 Key Components of a Bicycle Frame
The main function of the frame is to provide a central support structure to which other key components are attached. Its position as the central structure results in it absorbing the bulk of the strains and stresses caused by operation of the bicycle. As a result of this there are numerous structural requirements and thus by extension mechanical properties that a frame design must meet. In addition to these requirements, there are various specific customer needs that must be meet to ensure a desirable product is produced. Via research into cyclo-cross forums and racing magazines a list of needs was developed and is displayed in Table 1 [1-3].
Table 1 Customer Needs and Corresponding Metrics Needs Metric Unit Light weight Density Kg/m3 Absorb impact from jumps Toughness MPa Resist buckling during racing Stiffness GPa Minimise vibrations during riding Dampening GPa Wont failure under high impact Yield Strength MPa Corrosion resistant Corrosion Resistance Binary Long Lasting/ Durable Fatigue Strength MPa Relaxed geometry angles Frame dimensions Binary Flattened top tube for easier carrying Top Tube Shape Binary
Various needs outlined in Table 1 relate to the structural capability of the frame and as such a simple structural analysis is required to determine the required mechanical properties. Figure 2 displays a simple free body diagram of frame. During operation the ride imparts a load on the seat, handlebars and pedals resulting in loads transferring to the seat and head tubes and the bottom bracket. To balance these loads reactions forces are produced at the wheel hubs. As a result of these loads, internal shear forces and moments are generated in the frame members. Additionally the riders pedalling action creates a torque moment resulting in torsion shear stresses. These combined stresses create an overall stress state in each of the tubes. To avoid yielding and hence failure, the frame material must be selected such that the yield stress is greater than the equivalent Von Mises Stress in any section of the frame [4, 5].
Figure 2 Free Body Diagram Bicycle Frame
The nature of the compressive loading on the tubular members of the truss frame ensures failure due to buckling is a possibility. Thus the frame must have sufficient stiffness to ensure the critical buckling load is not reached [5]. Whilst preventing buckling in the frame is critical, the nature of cyclo-cross means the frame must be also be able to absorb the impact energy of various jumping and direction changing actions. To ensure good handle ability and a smooth ride the frame must be able to absorb the impact of these actions along with the forces created by pedalling. To ensure the frame does not experience severe crack initiation, growth and fracture the material should have sufficient toughness to enable impact energy to be absorbed and fracture toughness to ensure crack initiation and growth is prevented [6].
The repetitive pedalling action makes the lower joints in frame being susceptible to fatigue failure via the formation of fatigue cracks from cyclic loading. To ensure this failure doesnβt occur the frame material must have sufficient fatigue strength to enable long lasting use under stressful conditions.
Whilst the yield strength dictates when a material will begin to plastically deform, the hardness of the material also indicated its resistance to plastic deformation (shape changes) along with the wear resistance [6]. Bicycle frames must have an adequate hardness to resist plastic deformation at stress concentrators such as connections to other parts arms.
Corrosion resistance is critical in the sport of cyclo-cross due to the varying operating conditions and the importance of bicycle presentation in gaining critical sponsorship. Localised corrosion of frame joints can result in pitting and corrosion cracking, both of which can cause failure of the frame [7]. Thus the frame material must display a good resistance to corrosion either naturally or through further processing such as galvanising or coating.
Table 2 displays a summary of the required mechanical properties for the frame. These requirements form the basis of the ranking matrix used to select the ideal material.
5
Table 2 Material Property Requirements and Targets Need Material Property (Metric) Target Value High wear resistance Hardness (HV) >100 Resist crack formation and propagation Fracture Toughness (MPa.m1/2) >20 Impact energy absorption Toughness (Ut) Flexibility Elongation at break (%) >13 Light weight Density (kg/m3) <5000 Good stiffness in axial directions Youngs Modulus (GPa) >90 Good Stiffness under shearing forces Shear Modulus (GPa) >40 Wonβt yield when stressed Yield strength (MPa) >300 Long mean time between failure under cyclic loading Fatigue Strength (MPa) >80 Wonβt corrode in various environments Corrosion resistance Yes
For benchmarking purposes existing cyclo-cross franes were researched to grasp the materials commonly used for the frame [1-3, 7, 8]. It was observed that four different materials were consistently used depending on the specific requirements of the frame. Steels such as Chrome-Moly ASTM 4031 or Reynolds 725 Heat treated Chrome-Moly are commonly used for cheaper lower end bicycles [7]. Aluminium alloy frames are by far the most common, with the vast majority of middle to low end frames being constructed from Series 6XXX alloys, whilst higher end models are utilising series 7XXX alloys [3]. In particular AA 6061 T4 and AA 7075 are commonly used for middle to low end and high-end frames respectively [2-3, 7]. Carbon Fibre Reinforced Polymers is the most popular choice for high performance cyclo-cross bicycles but come attached with the largest price tag. The range in quality is quite large with lower end frames utilising less intricate laminate orientations and thus suffering in terms of strength and flexibility. The fourth material is Titanium alloy, which is found in the high end performance bicycles. The most commonly used alloys are Ti-3Al-2.5V and Ti-6Al-4V, with various companies experimenting with new manufacturing methods for frames made from these alloys [8].
Table 3 provides a summary of the advantages and disadvantages of the identified materials. Table 3 Comparison of common cyclo-cross bicycle frame materials [2-6, 10,11] Material Type Advantages Disadvantages
Aluminium Alloys
AA6061 AA7075
Low density, Good wear resistance and strength, Good corrosion resistance, Good price
Lower wear resistance and strength, low fatigue strength, lower strength. Lower hardness and ductility
Steel HSLA, Chrome-Moly
High strength, ductile, good fracture toughness and wear resistance, high fatigue strength, low cost, good reliability and damages can be easily repaired.
Higher density, fracture toughness -wear resistance trade off,
Titanium Alloys
Ti-6Al-4V, Ti-3Al-2.5V
Highest strength:weight ratio, good wear resistance, corrosion resistance, and stiffness, good dampening, good durability, Can be work hardened, good elongation and fatigue strength. Excellent tensile strength
Very high cost, manufacturing and machining difficulties, primary processing is expensive, Higher density than CRFP and Aluminium. Lower Youngβs modulus than steel
Composites CRFP
Very light weight, high stiffness, medium strength, medium wear resistance, can be formed into intricate shapes, Will not corrode, good dampening due to lower density
Isotropic, low fatigue endurance limit, very high cost, lower strength, anisotropic nature, very low elongation and hardness
Ceramic materials were ruled out due to their high mass, lower fracture toughness and low toughness when in tension and low thermal conductivity. Magnesium alloys were not considered due to the low strength and stiffness and poor corrosion resistance, whilst natural materials such as wood were discounted due to the limited production methods and inferior mechanical properties.
6
1.2. Material Selection Process The metrics and targets outlined in Table 2 were used to screen the material data bases enabling candidates that did not meet the requirements to be eliminated. Simple material indices were then derived using the objective and constraints outlined below.
1. Function: Provide central structural stability 2. Objective: Minimise mass 3. Constraints: Must meet yield strength, stiffness, hardness and fatigue strength criteria
The derivation of the material index, used to rank the materials based on strength and mass, is shown below. Initially the objective equation of minimising mass of each frame tubes is formed ππ= π΄π΄π΄π΄π΄π΄ Where m is the mass of the tube, A is the cross sectional area, l is the length and Ο is the density. As the length is most likely fixed to meet the riders needs, the mass can be reduced by reducing the cross sectional area [12, 13, 14]. However, a constrain exists whereby the area must be sufficient to carry the load F with out failing ππππβ€ πΉπΉ π΄π΄ The area is common to both equations and can be therefore be eliminated, ππβ₯ πΉπΉπ΄π΄π΄π΄ ππππ Inverting the final material property section allows the material index M to be maximised to find the lightest material that can safely support the load F. ππ=ππππ π΄π΄ This process was followed to develop materials indexes for the stiffness, and hardness criteria. The CES Edu pack software was used to generate a range of graphs using the material indices with the density occupying the X axis and the material properties constraints on the Y axis. The aim of the selection phase is to minimise the density whilst meeting the property requirements.
Figures 3-5 display the output from ranking process using CES material databases. The initial material property targets were used to identify the suitable materials. The material indices were then applied to rank the potential materials.
7
Figure 3 Upper left) Yield Strength v Density, Upper right) Compressive Strength v Density, Lower Left) Youngβs Modulus v Density, Lower Right) Hardness v Density
Figure 3 displays the material plots for four of the key material properties. It can be seen that after the initial design constraints are placed only three material groups remain, Metals and Alloys, Composites and Technical Ceramics. Guidelines representing the required materials index slope have been placed on each of the plots. As the lines progress up the plot the materials positioned above them are seen to be the best performers in for that materials index. It is noted that the technical ceramics are seen to perform quite well particularly in the compressive strength and hardness categories. CFRP are observed as the next best material closely followed by titanium alloys with the latter outperforming the former in the hardness category. This indicates the titanium alloys would be more durable in their given application, an important parameter when constructing a cyclo-cross bicycle.
The results of these plots indicate that both CFRP and Titanium alloys provide good strength to weight ratios and are capable of meeting the design strength requirements for the frame. The high performance of the technical ceramics is expected given their preferential use in compressive and high wearing applications. As will be seen later, whilst the technical ceramics outperform in they simply do not have the formability or fracture toughness to be considered for a frame material.
Figure 4 displays the material plots for 4 other key material properties. These plots focus on the materials ability to absorb energy during impact and to prevent the formation and propagation of cracks. Overall the titanium alloys are seen to perform the best in these categories, particular with respect to elongation and fatigue strength. This again indicates the high durability of a titanium alloy whilst also highlighting its good energy absorption and ability to prevent crack growth and propagation. The latter property is very important in ensuring catastrophic failure does not occur.
The CFRP, whilst still offering high performance for its density, does not perform as well as the titanium alloys in these categories. Other metals and alloys such as low alloy steel are seen to perform quite well but with the higher density are not considered the best option for a lightweight cyclo-cross frame.
8
Figure 4 Upper left) Fracture Toughness v Density, Upper right) Elongation v Density, Lower Left) Fatigue Strength v Density, Lower Right) Tensile Strength v Density Whilst the importance of the material properties for the frame cannot be understated, consideration must also be given to factors such as the materials environment impact and the ease of processing. In modern times the emphasis placed on producing an environmentally friendly product via minimal CO2 emissions during production, low embodied energy and high recyclability is ever increasing. Additionally to ensure the frame can be manufactured and sold profitably the material must be able to be processed easily with a low susceptibility to defects or other impairments on quality.
Figure 5 displays various environmental and economic properties. Both CFRP and titanium alloys meet the requirement of being recyclable; whilst titanium alloys are observed to have a lower cost, lower embodied energy and a slightly lower CO2 footprint.
9
Figure 5 Upper left) CO2 Footprint v Density, Upper right) Recyclable v Density, Lower Left) Price v Density, Lower Right) Embodied Energy v Density
In understanding the processing capability of CFRP and titanium alloys, thought must be given to the specific requirements of the frame. To meet the design requirements the material must be able to be formed into tubular shapes of varying thicknesses and shapes. CFRP has great mould ability and can be shaped very intricately but production lines are capital intensive and tight control is needed to ensure product quality is maintained [10]. Likewise due to the difficult in the primary processing of titanium ores, the cost of production of titanium alloys is high. The mouldabilty is not has good as CFRP but the ability to butt frame tubes and the ability to work harden means very high strength to weight ratios can be obtained [10]. Additionally the good toughness and elongation of titanium alloys enables a flexible but sufficiently strong and stiff frame to be manufactured. These critical factors produce a more durable and accommodating frame with high handle ability, all of which are important factors in cyclo-cross racing.
The findings from the material selection process indicate that titanium alloys are the best choice for a cyclo-cross bicycle frame. The very high strength to weight ratio combined with good stiffness, toughness and elongation ensures the frame can withstand the rigours of the racing environment whilst also providing a responsive, durable ride. In particular titanium alloys good dampening ability ensures the shock loadings from impacts during race are not transferred directly to the rider. Titanium alloys good fatigue strength and fracture toughness ensure a long lasting and durable frame can be built with the chance of catastrophic failure via crack formation and growth minimised. Additionally, whilst still relatively high, the environmental impact and economic properties of titanium alloy production was lower than the closest rival CFRP.
10
To determine a specific titanium alloy, CES software plots were again utilised but with the material database limited to titanium alloys. The critical material properties, yield strength and stiffness, were used to select the ideal material. Figure 6 displays the respective material plots. As can be seen, there is a vast range of alloys available. Whilst maximising the strength to weight ratio is important, there is a limit to the required strength and thus with the alloys displaying similar densities it was determined that a cheaper but sufficiently strong alloy would suffice. Additionally benchmarking research revealed multiple titanium alloy frames constructed from Grade 9 Ti-3Al2.5V. After reviewing the performance of this particular alloy is was selected as the material for which the frame would be made from.
Figure 6 Upper left) Youngβs Modulus v Density, Upper right) Price v Density, Lower Left) Yield Strength v Density, Lower Right) Compressive Strength v Density
1.3. Ti-3Al-2.5V Ti-3Al-2.5V is a commercial Ξ±/Ξ² titanium alloy commonly used in the aerospace industry for aircraft and engine hydraulic system tubing. The alloy also finds use in golf club shaft and highend bicycle frames [CES, other]. Table 3 displays the chemical composition of the Grade 9 Ti-3Al2.5V titanium alloy selected for the cyclo-cross frame [15].
Table 4 Chemical composition of Ti-3Al-2.5V [15] Ti (%) Al (%) V (%) C (%) Fe (%) N (%) Bal 2.5-3.5 2-3 0-0.08 0-0.25 0-0.02
11
Figure 7 displays the typical phase diagram for titanium with alloying element content on the x axis. At room temperature titanium has a HCP structure known as Ξ±. At approximately 882.5Β°C an allotropic transformation occurs whereby the microstructure changes from the Ξ± HCP structure to a BCC Ξ² phase [polymear]. This Ξ² phase is then stable until melting occurs at approximately 1678 Β°C. The presence of this transition enables titanium alloys to be formed with Ξ±, Ξ² or mixed Ξ± + Ξ² microstructures [polymear]. The ability to create these different microstructures derives from the stabilising effects of different alloying elements. Elements such as Al, O and N all dissolve preferentially in the Ξ± field, acting to expand it and raise the Ξ±/Ξ² transition temperature stabilising the field in the process. Conversely elements that depress the Ξ±/Ξ² transition temperature act to stabilise the Ξ² field and can be classified into two groups. 1. Favouring Ξ²-Isomorphous type 2. Favouring Ξ²-Eutectoid type
Figure 7 Titanium Alloy Phase Diagram In the case of Ti-3Al-2.5V the vanadium addition acts as a Ξ² stabiliser whilst the aluminium acts as an Ξ± stabiliser [14]. The Ξ± + Ξ² microstructure in the alloy consists predominantly of equiaxed Ξ± grains with small regions of transformed Ξ² positioned at the Ξ± grain boundaries. This structure is created via quenching from the Ξ² phase and then annealing in the Ξ± + Ξ² region followed by slow cooling. During cooling the Ξ± phase forms as equiaxed grains in amongst the Ξ² matrix. As further cooling occurs the regions of Ξ² transform into laths of widmanstatten Ξ± [14]. This microstructure is displayed in Figure 8 [14].
Figure 8 Ti-3Al-2.5V alloy annealed in the Ξ± + Ξ² region. The left image displays the equixed Ξ± grains and transformed Ξ². The right image shows a close up of the transformed Ξ² as widmanstatten Ξ±
12
The presence of the equiaxed Ξ± grains along with the widmanstatten transformed Ξ² results in increased yield strength, fracture toughness and fatigue strength. The mixed microstructure provides a high resistance to the initiation of fatigue cracks and the high number of interfaces and grain boundaries results in very slow propagation of cracks [14]. The increased number of grain boundaries provides a high strength in accordance with the Hall Petch Equation [14,18]. The presence of equiaxed Ξ± grains amongst the transformed Ξ² ensures the alloy has a good level of ductility. In fully widmanstatten alloys a basket weave structure is created resulting in reduced ductility as a result of the acicular widmanstatten laths. Whilst this is beneficial for fracture toughness, fatigue life and ductility suffer due to the lack of deformable Ξ± grains [14,16].
Solid solution strengthening is present in the alloy via the addition of Al and V. In solid solution these elements substitute for titanium in the matrix creating a distortion in lattice [14]. This distortion creates a stress field in the lattice, which acts to slow down dislocation movement thus increasing the strength of the material [16].
Typically the Ti-3Al-2.5V alloy is used in the mill annealed condition, with heat treatment taking place in the Ξ± + Ξ² region. Table 4 displays the typical material properties of a grade 9 Ti-3Al-2.5V alloy resulting from approximately as 30% volume of equiaxed Ξ± grains amongst the transformed Ξ².
Table 5 Material Properties Property Unit Value Youngβs Modulus GPa 107 Yield Strength MPa 550 Tensile Strength MPa 620-750 Elongation % 15-20 Compressive Strength MPa 580-744 Modulus of Rupture MPa 531-682 Shear Modulus GPa 33-36 Poissonβs Ratio 0.35 Hardness HV 103-113 Fatigue Strength MPa 363-432 Fracture Toughness MPa.m1/2 70-75 Thermal Expansion Strain/C 7.9-8.1 Thermal Conductivity W/m-C 7.5
13
2. Manufacturing Process This section details the selection methods used to determine the most applicable process for manufacturing the bicycle frame. The selected process is has been outlined in detail with a corresponding flow chart developed for clarity.
2.1. Process Selection For a particular product and material a range of manufacturing process options exist. Selecting the best process requires an understanding of the required product properties and how each option affects these final properties. In case of the bicycle frame, initial background research on the primary processing of titanium alloys was completed to understand the formation of cast and wrought alloys. Further research was conducted to form a list of potential processes for making either a one piece of multi piece frame. The processes were then ranked based on their advantages and disadvantages to determine the best option. 2.1.1. Primary Processing: Kroll Process The production of titanium cast ingots is completed via the Kroll Process, which is displayed in Figure 9 [14,15]. Initially a titanium dioxide is mixed with coke and charged in a chlorinator. This results in a reaction between the ore and the chlorine forming titanium tetrachloride liquid. The liquid is purified via fraction distillation before being mixed with powered magnesium and sealed in a vacuum container void of oxygen and hydrogen. Through the application heat, the magnesium powder reacts with the chlorine to produce magnesium chloride, leaving behind pure titanium in a sponge like format. To increase the density, the sponge is compressed by hydraulic presses to form titanium βcompactsβ [15]. The compacts are joined together via tungsten inert gas (TIG) welding to form a consumable electrode.
To form the titanium alloy the electrodes are melted down in a copper lined vacuum electric arc furnace to form molten titanium. The alloying additions, in this case Al and V, are then added and agitation used to create a consistent mix. Once full mixing has occurred the molten mixture is then allowed to cool and solidify in the furnace. After solidification the ingot is removed along with the attached copper lining, which is then removed via a lathe. The product at this stage is in the form of a cast ingot. While this ingot can be immediately utilised for further processing such as die casting, it is more common for the ingot to be reduced in size via forging in preparation for further shaping processes such as extrusion and pilgering [15, 18]. During this stage the ingot is continuously hammered to reduce the size and then annealed to remove work hardening effects and re-introduce a good level of ductility. The process is conducted in a vacuum environment to ensure oxygen and nitrogen do not react with the titanium surface resulting in the formation of a hard and brittle Ξ± case phase and to minimise oxidation scaling [19]. The forging process is repeated until a wrought 8 inch bar is formed.
Figure 9 Kroll process for titanium sponge production
2.1.2. Primary Shaping Options A bicycle frame can be constructed either as one whole part or via the formation of a number of different dimensioned tubes, which are initially formed from billets before being shaped into the required geometry. In the case of the latter option a joining process is required to complete the frame. To determine the processes that could potentially be used to manufacture the frame the following generic screening matrix was developed based on the ideal needs for the final frame.
Table 6 Screening matrix for ranking potential manufacturing processes Need Metric Strong and ductile material Equiaxed Ξ± grains + widmanstatten transformed Ξ²
Can produce the frame economically Low unit cost, low capital and tooling costs
High quality Free of internal defects, minimal stress concentrations/weak spots Minimise waste Low scrap fraction Meet tight required dimensions Tight dimensional and geometrical tolerances
Aesthetically pleasing Quality, defect free surface finish Minimise variability in product Tight process control Automated batch processing with minimal production delays Incorporate well with upstream and downstream processing
Using this methodology background research was conducted into the possible titanium processing methods. These methods were shortlisted to 4 processes and are briefly outlined below.
Seamless Extruded Tube Forming Billets are heated into the range of 850-1000ΛC and then pierced onto a mandrel whilst pressing against a series of rollers or extruded through a die onto a mandrel to form a tube hollow. The former option requires less lubrication and avoids the cost of expensive extrusion dies [15, 18].
The tube hollow then reduced in wall thickness and diameter by either pilgering, tube rolling or cold drawing [17, 18]. During pilgering the tube hollow is supported and rotated by a mandrel as it is pushed through an oscillating mill case. The mill case contains two ring dies, which due to the mill cases motion; roll back and forth over the rotating tube reducing its diameter and thickness. Tube rolling utilises one to two sets of rolls with constant section grooves on the circumference to roll out the tube hollow. Cold drawing reduces the diameter of the hollow by pulling it through a die that is smaller than the tube. The formed tube then goes through a forming stage to create the desired shape for the frame. The possible forming processes are discussed later in this section.
The main advantage of seamless tube forming is defect free surface finish as opposed to welded tube whereby the weld line poses as a stress concentrator and crack initiation site. Additionally seamless extrusion enables a tight control over the final microstructure when combined with vacuum annealing.
The advantages of the pilgering process include high reduction ratios, tight control over the grain structure, reduction in material eccentricity and defect free reductions.
Welded Tube Forming Hot or cold rolled titanium strip is formed into a circular cross section via a series of contoured rolls [18]. A gas tungsten arc positioned immediately after the rolls then welds the two free edges together as the strip travels at speeds up to 7.6m/min [18]. The welded zone is shielded by an inert gas to prevent oxidation of the weld and the heat affected zone. The tube then undergoes a similar
15
reduction process as the seamless tube with the cold drawing being the most favourable as it enables the weld bead to be deformed and smoothened [18].
A key disadvantage of welded tube forming is the orientation of the grain structure. During rolling the grains are elongated in the rolling direction. When the strip is bent into a circular cross section the grains at the weld may be orientated differently resulting in a possible weak spot or stress concentration site [15].
Casting Previously cast ingots are heated to above the Tm to form a molten Ti-Alloy. The molten metal can then be cast in into either an investment mould or a rammed graphite mould, with the latter being similar to the sand casting method.
Casting titanium suffers from a range of issues which prevent it being a leading manufacturing method for the mass production of tubes. Its high chemical reactivity means inert mould surfaces such as lost wax ceramic shell systems. Additionally titanium has a high affinity for atmospheric gases, which can result in the formation of gas pores and internal defects in the casted structure. Finally the flow properties of molten titanium are severely reduced and this results in issues with mould filling. Poor mould filling results in porous sections which form the sites of crack formation in as cast structures.
3D Printing 3D printing can be used to manufacture the bike in one complete part. Initially the titanium alloy is crushed via a mill into an extremely fine powder. A 3D CAD model is then constructed and uploaded to the printing machines software with minor adjustments to the geometry and positioning. The structure is then divided into layers, which are downloaded to the direct metal laser sintering machine (DMLS). The DMLS machine then fires a powerful optical laser into the building chamber where a specialised platform dispenses the powdered titanium alloy over a coating blade to form a respective layer. The titanium powder is fused into a solid form via local melting created by the focus of the laser beam. The bicycle frame is constructed layer by layer with the geometrical features controlled by the 3D CAD model.
Table 7 summarises the various processing options for manufacturing the tubular sections.
Table 7 Tube Manufacturing Processes Process Advantages Disadvantages Seamless Tubing Low scrap fraction, Fine grain structure, minimal need for downstream machining, fast cycle times, cheaper tools, good surface finish, narrow tolerances, Uniform shape Requires lubrication between rolls and tube, Surface defects can arise, Capital costs high, pickling stages needed. Welded Tubing More cost effective than seamless tubes, Better surface quality apart from weld. Narrow tolerances and low scrap fraction Grain orientation may cause premature failure. Weld seam is a stress riser. Casting Cost reduction, equal in strength to wrought alloys, equal or higher fracture toughness. Production rates higher Shrinkage, Gas porosity, cold shuts. Surface defects present. Dimensional tolerances bad. Lower fatigue strength, poor surface finish, lower ductility, lower fatigue strength, high chemical reactivity with mould, poor molten flow properties 3D Printing High speed process with minimal preprocessing. Complex structures can be formed. Very low waste, very tight tolerances, Poorer surface finish, reduced strength, increased amount of defects, higher porosity. Size capabilities, reduction in fatigue strength.
16
The tubing formed form the primary shaping process can undergo further processing to shape the tube into its desired form.
Tube Butting Tube butting is a process used to create different wall thickness along the length of the tube. The tube can either but internally or externally butted depending on the whether the inner or outer diameter is changed [15].
During internal butting a mandrel press pushes the tube hollow through a die pressing it down onto the mandrel. The dimensions of the die determine the outside diameter of the tube which is kept constant, whilst the shape of the mandrel determines the inner diameter and wall thickness. As the material is pressed onto the mandrel material is pushed towards the sections where the mandrel diameter is reduced.
To remove the mandrel from within the tube, the outside of the tube is put though a reeling process whereby it is passed between two angled rollers. These rollers act to increase the diameter whiles having a negligible effect on the wall thickness. To reobtain the required outer diameter the tube is passed through another extrusion die.
Hydro Forming The tube is placed into a forming press that contains a cavity matching the required tube shape. The hydropress is closed and sealing rods inserted into either end of the tube. High pressure water is injected into the tube whilst a force is applied to the sealing rods to keep the tube in place. The pressure from the water yields the tube and it deforms to fill the shape of the die.
Table 8 outlines the possible methods for options for forming the specific tube sections.
Table 8 Tube Shaping Processes Process Advantages Disadvantages Buttering High dimensional tolerances, stiffer and stronger, good surface finish, can create multiple wall thicknesses in one tube, high fatigue strength Requires annealing to relieve cold working Hydroforming Cost effective, creates light weight parts, increased strength high quality surface finish, Lower thickness tolerances, Lower strength and fatigue strength, requires high ductility to avoid cracking
Based on the information presented, the combination of seamless tube formation from forged alloy billets and tube buttering was selected as the overall manufacturing process. This process was selected for the following reasons;
β’ Tight control over final microstructure o The process enables a grain structure consisting of equiaxed alpha grains and transformed beta. This grain microstructure provides the required mechanical properties needed for the frame design.
β’ Narrow dimensional and geometrical tolerancing o The process enables tight diameter and thickness tolerances along with good concentricity. Other processes such as casting offered poor tolerancing. o β’ Minimal waste and low scrap fraction o There is minimal material loss during each of the processing stages resulting in low material costs
17
β’ Quality surface finish o Minimal surface defects β’ Minimal need to downstream machining o Good surface finish reduces the need for further machining and treatments β’ Good mechanical properties o High yield strength, fatigue strength and ductility β’ Process variables can be monitored for tight control o Main variables of temperature, annealing atmosphere, strain rate and pressure can all be monitored and kept within a tight band. 2.2. Seamless Tube Manufacturing The seamless tube manufacturing process begins with the Ti-3Al-2.5V billet created through the earlier forging process. This billet is put through various processes to produce the final tube. To create all the components for the frame the process is completed 6 times. 2.2.1. Roll Piercing The billets formed during the previous forging stage are placed in a vacuum induction furnace and heated to the range of 850-1000 degrees Celsius and held for approximately 30 minutes [16]. At high temperatures titanium is high reactive. To prevent the formation of scale and surface defects the heating environment is void of hydrogen, oxygen and nitrogen. The heat acts to relieve any residual stresses in the billet caused by the forging stages [15, 16].
The billet is removed from the furnace and immediately fed into the roll piercing set up. The billet is then passed into the piercing chamber where a hydraulic press drives it into a roll setup consisting of two lubricated conical shaped roles, a lubricated piercing mandrel and shoes to support the billet. The rolls compress, rotate and pull the billet forward simultaneously [19]. As the compression occur centre of the billet is placed under tension and the rotation results in a cavity opening up in the centre. The billet is then pierced by the plug and the presence of the cavity enables further piercing and material flow over the mandrel forming a tube hollow. The hollow is elongated by the compressive force of the rollers to form the optimal length and diameter.
The tube hollow is then immediately immersed in a bath of hydrochloric acid to pickle off the superfluous outer layer and remove any brittle alpha case that may have formed. Finally the tube is cut to length for the next forming stage. At this point one tube hollow provides the basis for the all the different sized sections of the frame.
Figure 10 Billet piercing process
18
2.2.2. Vacuum Annealing During the piercing and rolling, the titanium alloy significantly work hardens resulting in a strong but less ductile material. To remove residual stresses formed by this work hardening stage the tube hollow is heat treated. The tube is first weighed and the dimensions are verified using radars. The tube is then mill annealed at approximately 700C [16] in a vacuum induction furnace to remove residual stresses created during the drawing process. The vacuum environment is required to prevent the pickup of oxygen and nitrogen, which result in the formation of the brittle surface phase known as Ξ± case [16]. Removing the alpha case prevents the formation of surface cracks during further processing, improving the machinability and increases the tool life. Additionally, preventing contact with oxygen helps to minimise surface scaling. The tube hollow is then pickled in a bath of HCl to remove any oxidation layers or layers of alpha case as this is can cause embrittlement and fracture during the pilgering process.
The annealing heat treatment acts to stabilise the tube shape and eliminates strength conditions, such as the Bauschinger effect, created during the roll piercing stage. Annealing also greatly improves the ductility enabling the tube to be processed through the pilger machine without brittle failure occurring. As the annealing can cause the tube to bend or bow, it is creep straightened at a temperature above 525C. Straightening at this temperature avoids the issue of spring back which is heightened in titanium alloys [16].
Figure 11 Vacuum annealing furnace
2.2.3. Pilgering At this stage of the process the tube hollows all contain the same wall thickness and inner and outer diameters. To create the different tube dimensions used in forming the bicycle frame the tube hollows undergo a series of cold working via a Pilger machine.
The cold pilgering process uses ring dies and a tapered mandrel to reduce the cross section of the tube hollow [15, 18]. Initially the mandrel, which is tapered in the direction of rolling, is inserted into the tube and fixed into position. The mandrel and tube are then passed into the mill saddle which contains the ring dies. As the tube is pressed forward and it is rotated by mandrel. At the same time the mill saddle is driven back and forth creating an oscillating rotary motion in the ring dies. During the forward and backward strokes of the mill saddle the ring dies apply a deforming load on the tube, reducing both the wall thickness and diameter.
Each tube goes through the pilgering processes several times as some passes focus on reducing the tubes diameter whilst others emphasize reduction of the wall thickness. The number of passes required for each tube is dependent on the groove design of the ring dies and the mandrel shape. The repeated pilgering is performed as part of a sequence that includes pickling in baths of hydrochloric acid (HCl) to remove oxidation scale, annealing in oxygen free environments and straightening through tapered angled rollers, followed by ultrasonic testing to verify the tubes dimensions. This sequence continues until the nominated tube size and material grain structure are achieved [15].
19
Once the wall thickness, diameter and grain structure are with in specification the tube is pickled again in a HCl bath to remove scale and debri before a final vacuum anneal occurs. This final annealing stage is used to relieve any residual stresses and create a balance between strength and ductility [14]. After the final anneal the tube is acid etched again to remove any oxidation or alpha case that may have formed [sevens].
Figure 12 Tube pilgering process showcasing the ring dies and mill casing
2.2.4. Final Heat Treatment To impart the required mechanical properties into the tube a solution treatment + quenching + aging stage is used. The tube is first heated to ~25-85C below the Beta transus temperature for approximately 30 minutes to create a solid solution of alloying elements in a matrix consisting predominantly Ξ² with a smaller percentage of Ξ±. The tube is then quickly removed from the furnace and water quenched to trap the Ξ² in an unstable state whilst also forming a HCC martensitic Ξ±β phase. The tube is then aged in a vacuum furnace at approximately 425-600ΛC to transform the unstable Ξ² into widmantstatten lath Ξ± and fine precipitates of Ξ² bucleat on the boundaries of the Ξ±β laths. This microstructure results in an increased yield and compressive strength whilst also maintaining good ductility and enhancing the fatigue strength and fracture toughness. The tube is then pickled in a HCl acid bath to remove any scale or alpha case.
2.2.5. Tube Butting
With the tube now formed to the required dimensions and grain structure, the process of internal butting is utilised to create a variable thickness in the tube. A mandrel of with a reduced diameter at the end sections is inserted in to the tube. A hydraulic press then pushes the tube through an extrusion die, sinking the material down onto the mandrel [15]. The outer diameter is dictated by the dimensions of the die as it smooths the outer surface of the pipe onto the mandrel. The inner diameter and wall thickness are therefore dictated by the profile of the mandrel. The end result is the formation of a thinner tube section where the diameter of the mandrel is greatest and thicker tube sections at the smallest mandrel diameter locations. This is done to enable more material at the highest stressed areas of the, the connections. These areas require extra wall thickness to ensure yielding and failure does not occur along [15]. To ensure the frame weight is minimised the tubes are butted to decrease the thickness in regions of lower stress.
After the butting process is complete the inner diameter of the tube is greater at the end sections effectively trapping the mandrel inside the tube. To remove the mandrel, the tube is passed between angled rotating rollers, which increase the diameter while having a negligible effect on the wall thickness. The mandrel is removed and the tube outside diameter is resized by pushing it through an extrusion die once more. The tube is then cut to size using a band saw and any cabling holes are drilled via a drill press.
Figure 13 Tube butting and reeling process
20
The following process flow chart outlines the entire process for create one tube section of the frame. The process would be tailored to meet the dimensional requirements of each tube section.
The billets are heated in a vacuum furnace to just below the Beta transition temperature to create a solid solution of the alloying elements. The heating reduces the strength of material thus decreasing the flow stress enabling hot forming to be conducted with lower energy.
The heated billets are pressed into a piercing mandrel and pushed through a series of rollers to form the tube hollow. Work hardening occurs resulting in an increase in strength.
HCl The tube hollow is pickled in a bath of HCl to remove the superflous outer layer
The tube is vacuum annealed at 85C below the Ξ² transus for approximately 30- minutes to remove residual stresses and increase ductility for further cold working. Vacuum environment is needed to prevent reaction with oxygen and nitrogen which results in the formation of alpha case
The tube is creep straightened at ~ 550C through a series of rollers to remove bends that may have occurred during annealing
The tube passes through the pilgering machine to reduce the diameters and wall thickness. After each pilgering stage the tube is annealed, pickled and straightened. The process is finished when the correct grain structure and tube dimensions are formed.
HCl The tube hollow is pickled in a bath of HCl to remove any oxidation layers or debri
The tube is solution treated just below the Beta transus for 30 minutes to create a solid solution. Vacuum environment is needed to prevent reaction with oxygen and nitrogen which results in the formation of alpha case
The tube is butted to create a increased wall thickness at the tube ends
The tube is aged at approximately 425-600C to decompose the unstable Ξ² and the hexagonal Ξ±β to form equiaxed Ξ± and widmanstatten Ξ± laths. An increase in toughness and ductility is created. At a temperature above 425C avoid brittle omega phase formation.
HCl The tube hollow is pickled in a bath of HCl to removeany oxidation layers, debri and alpha case
The tube is cooled rapidly to form a unstable Beta phase and heaxgonal alpha β structure. A delay of time from the furnace of less than 7 seconds is required to ensure unstable beta is formed
21
2.3. Combined Frame The process described above is used to manufacture individual tubes. As shown in Figure 14 a bicycle frame consists of 8 different tube sections. Based on the standard 56.5cm sizing the total length of tube hollow required is approximately 280cm. The required dimensions of each tube are outlined in Table 9 and Figure 14 [3]. Table 9 Tube dimensions (mm) Top Tube Down Tube Seat Tube Seat Stays Chain Stays Head Tube Bottom Bracket Do tw Do tw D o tw Do tw Do tw Do tw Do tw 0.7/0.5/0.7 57 0.8/0.5/0.8 0.8/0.5/0.8 0.6/0.4/0.6 0.6/0.4/0.8 1.6/2.3 1/0.55
Figure 14 Tube lengths for standard size frame
To manufacture all the frame components a single tube hollow would be produced from the forged billet. Cutting this hollow into sections would enable the different sized tubes to be formed in the pilgering and butting stages. These process would need to be completed a minimum of 6 times to create all the required tubes.
57cm
64 cm
55 cm
10 cm
47 cm
40 cm
22
3. Operating Costs This section provides the calculated production costs of the manufacturing process.
3.1. Key Assumptions The following assumptions were used when determining the manufacturing costs;
β’ The 1000m2 industrial factory is located in Perth, Australia β’ 1x 12hr/day operation shift would be utilised 5 days a week. β’ Scrap fraction for the process is approximately 0.1 for the process β’ The frame size is 56cm standard 3.2. Material Costs The advantage of the seamless tube forming process is the small amount of raw material needed to form the final product. The exact amount of lubricant and HCl required per frame is difficult to determine and as such an estimate is given. The price of Ti-3Al-2.5V billets were obtain in $/kg and the required billet mass determined from the dimensions in Figure 15.
Figure 15 Billet Dimensions Table 10 Material Costs Material Price Amount per frame Total Cost ($)/unit Ti-3Al-2.5 Forged Billet1 500 $/Billet 1 500 Lubricating Oil 12.84 $/L 1 12.84
3.3. Tooling Costs The tube forming process requires a number of different tools. The process does benefit from a long tooling life, which reduces the unit cost. The typical life of each tool was derived from research into test work done on typical life spans. In certain cases the tool life was not found through research and instead the value was estimated. In the case of the working rolls, the ability to repair the surface via grinding greatly increases the tool life. In the case of the HCl a single bath of 0.25m3 is needed. The life span of the HCl has been estimated based on the contimantion from scale formation. Table 11 Tool Costs and Life Tool Price ($/Tool) Tool Life (Units/Tool) Piercing Mandrel 1200 3202 Working Rolls 700 2000 Guide Shoes 500 800 Straightening Rollers 600 2000 Pilgering Mandrel 800 300 Pilgering Ring Dies 1200 500 Butting Mandrel 500 1000 Butting Die 900 400 Hydrochloric Acid3 74 500
1 http://www.alibaba.com/product-detail/price-per-pounds-for-Ti-3Al_60248562361.html 2 https://www.stahleisen.de/Portals/stahleisen/11-TT-Pehle_Originalversion.pdf 3 http://www.alibaba.com/product-detail/hydrochloric-acid-price-for industry_1021672202.html?spm=a2700.7724857.29.55.T4z11J
3.4. Capital Costs The capital cost makes up a significant portion of the overall production costs due to the number of components required. Were possible the price of the pieces of equipment have been obtained. The write off time has been estimated as 5 years. The load factor has been calculated by factoring in the available production time, and scheduled and unscheduled maintenance on a yearly basis. Given the number of individual pieces of equipment it was elected that shutdowns for scheduled maintenance would occur every 14 weeks. Each shut down would last 24 hours to recondition the equipment according to the planned maintenance instruction and service life agreements. Table 12 Operating Hours and Load Factor Costs and Life Item Value Operating Hours 3120 hrs/yr Scheduled Maintenance 576 hrs/yr Unscheduled Maintenance 200 hrs/yr Holiday Periods 120 hrs/yr
Total Available Time 2224 hrs/yr Load Factor 0.25 Write Off Time 8736 hrs Production Rate 0.7 units/hr
The production rate is calculated from the first billet heating until the end of butting and final inspection process. The most time consuming part of the process is the heat treatment sections. Table 13 Process Step Times Step Time (mins) Furnace Heating 20 Piercing 1 Rolling 3 Pickling 2 Annealing 20 Straightening 1 Pickling 2 Pilgering 5 Solution Treating 20 Quench 1 Aging 20 Pickling 5 Butting 5
Total 177 Table 14 outlines the processes capital costs. The Seamless tube rolling plant covers the Piercing Mill, rotary hearth furnace, plug mill, reeler mill, Reheating tunnel furnace, sizing mill, cooling bed and sprays, and rotary straightener. Capital costs were obtained for similar mills such as extrusion and cold drawing. Table 14 Capital Costs Item Price ($) Amount Total Cost ($) Seamless Tube Rolling Plant4 800000 1 800000 Pilgering Mill 75,000 1 75,000 Butting Mill 40,000 1 40,000 Work Benches5 300 4 1200 Storage Rack6 500 1 500 General Tools 600 1 600 First Aid Kit7 50 1 50 Fire Extinguisher7 150 3 450
4 http://esteelcorp.com/tube_pipe_mill/STEEL_SEAMLESS_TUBE_MILL/ES-11178/168_ES-11178.pdf 5 http://www.steelspan.com.au/products/workshop-storage-systems 6 http://www.redbackstoragesystems.com.au/ 7 http://www.boc.com.au/shop/en/au-boc-industrial-store/fire-extinguishers
24
3.5. Overhead Costs Table 15 outlines the overhead costs. Costs such as sales, marketing, maintenance, insurance, employee superannuation and management costs were not included in the calculations as they were considered to in depth for the analysis. Table 15 Capital Costs Item Price ($) Total Cost ($/hr) Energy8 32.6c/kWh 13.6 Factory Rent9 57000 p.a 25 Labour10 40/hr 120
3.6. Unit Cost Analysis The cost per unit for a given unit production rate and production volume is given by; ππππππππ πΆπΆπΆπΆπΆπΆππ=πΆπΆπ π =πΆπΆππππππππππππππππ+πΆπΆππππππππππππππππππ ππ +πΆπΆΜππππππππππππππ+πΆπΆΜππππππππβππππππ ππΜ The equation is expanded to account for further parameters
The unit cost was calculated as a function of production volume from 1 to 10000 units per year. The maximum value of 10000 was used to obtain the limit minimum unit cost. In practice given the proposed production rate and available production hours the maximum number of units that can be produced per year is 1500.
The following figure displays the cost per unit against yearly production volume.
Figure 16 Unit Cost v Production Volume
8 http://esteelcorp.com/tube_pipe_mill/STEEL_SEAMLESS_TUBE_MILL/ES-11178/168_ES-11178.pdf 9 http://www.realcommercial.com.au/for-lease/property-industrial+warehouse-in-wa/list1?includePropertiesWithin=includesurrounding&maxFloorArea=1%2c000&minFloorArea=700&source=refinements 10 http://www.payscale.com/research/AU/Location=Perth-Western-Australia/Salary
$0.00 $1,000.00 $2,000.00 $3,000.00 $4,000.00 $5,000.00 $6,000.00 $7,000.00 $8,000.00
1 10 100 1000 10000
Unit Cost ($/unit)
Production Volume (Units)
Unit Cost v Production Volume
25
Based on the simple analysis performed above and the calculated maximum number of units per year of 1500, the minimum number of frames per year needed to make the business viable would be approximately 150. The high influence of the tooling costs ensures small production is very expensive. However as the production volume increases the impact of the tooling on the overall costs is reduced and the unit cost hits a minimum at approximately 800 units.
3.7. Major Cost Factors
Marketing and Sales and Distribution Marketing and advertising consume a large chunk of a businessβs expenditure but due to their critical role in selling the product they cannot be ignored. The wages of marketing and sales and customer response staff must be covered along with forms of advertising. In particular the market for cyclo-cross bicycles is very competitive with newcomers to the sport relying on product reviews and advertisements to influence their decision.
Inventory and Distribution Costs Distribution hubs are needed to effectively ship products to the customers. Lead times need to be minimised to meet customers demand in the modern fast paced world. Contrary to that it is costly hold large amounts of stock in inventory.
Research & Development To develop new competitive products a company must engage in the research and development of new materials and or manufacturing methods. Successful companies constant reinvigorate and rebrand themselves to the market to keep their company brand fresh in the mind of consumers. To do this the business must conduct research into new frame materials or manufacturing processes. This includes partnerships with universities and the use of pilot plants for trials.
Business Systems In order for a companyβs management and administration processes to be effective and efficient, they need to be consistently reviewed and stream-lined. Poor, ineffective and/or inefficient processes and practices will inevitably lead to incidents. These could be safety, production or quality related amongst others. Ensure business systems are maintained and effective is paramount to keeping a tight production set point. Implementing, maintaining and review business systems costs money but if not kept in check the repercussions can result in the business going bankrupt.
Maintenance Costs In a metal forming process, the requirement of functioning machining is high. Malfunctioning equipment can produce bad quality products which can not only cause immediate issues through product recalls but can also damage the brand of the business. Additionally a broken down piece requires prompt fixing as lost production hours can rapidly increase the unit cost of the product. Unscheduled maintenance is not beneficial to maintaining a tight production set point and philosophy whereby routine planned maintenance inspections and services are completed in much more beneficial. Planning the work and working the plan is an effective way to avoid break downs. All of this costs money and its input into the unit cost via the maintenance budget.
26
4. Process and Quality Control 4.1. Process Variables The main process variables involved in the tube manufacturing are the applied pressure (or strain rate), the temperature and the time. Additionally factors such as the operating environment must be taken into account given the high reactivity of Titanium. 4.1.1. Temperature Like other metals, the temperature at which titanium alloys are hot worked has a direct influence on the required working forces, possible reduction ratios, reactions with the atmosphere and the developed microstructure. Consequently the working temperature has a direct influence on the internal microstructure and subsequently the mechanical properties of the material.
Like other metals titanium alloys plastically deform via the movement of dislocations along slip planes. However, in the case of titanium the number of slip planes present at lower temperatures is greatly reduced compared to metals such as aluminium. Consequently a high level of heating is required to minimise the required flow stress. The flow stress required to plastically deform titanium alloys decreases as the temperature is increased as shown by Figure 17 [18]. As the temperature is increased, dislocation movement is increased due to thermal activation. The dislocations are able to climb and travel along an increased number of slip planes [14]. Additionally a phenomena known as creep deformation occurs allowing increased deformation at lower applied forces.
Figure 17 Strength v Temperature for Ti-3Al-2.5V Initially the forged billet is heated to a temperature just below the Ξ±ο Ξ² transition temperature. The heating acts to dissolve Ti3Al precipitates to create a solid solution of the alloying elements Al and V in the Ξ±/Ξ² matrix. Removing precipitates is critical as they can effectively pin dislocations resulting in increases in the flow stress. The heating also serves to remove residual stresses caused by the previous forging stages. When present these residual stresses create strain in the matrix, impeding the movement of dislocations and thus increasing the flow stress. Additionally heating to just below the Ξ² tranus serves to create a microstructure of equiaxed Ξ± in a Ξ² matrix. The required flow stress for this microstructure is less than a structure consisting of acicular Ξ±. Care must be taken to avoid heating above the Ξ² tranus as subsequent deformation and cooling results in a reduction in tensile properties, in particular ductility, which can not be recovered by further heat treatment.
Above approximately 425ΛC titanium is reacts readily with oxygen and nitrogren in the atmosphere. The reactions result in the formation of oxidiation scale and the formation of a surface Ξ± phase called alpha case. This phase is very brittle and reduced the fatigue life of the material. As the highest stress point of a tube is located at the surface, the presence of brittle phases can induced fatigue cracks and result in brittle failure of the part. To prevent the formation of these phases, heat
27
treating must be conducted in a vacuum furnace. Additionally the vacuum aids to prevent hydrogen embrittlement, which occurs when hydrogen diffuses into the surface layers of the alloy.
The initial heat treatment of the billet reduces the flow stress as it is pierced and elongated into the tube hollow. Temperature is again involved in the heat treatment applied prior to cold working. Heat treating at just below the beta transus relives the stresses created via strain hardening during the deformation stage and increases the ductility of the tube. During straightening, a sufficient temperature of >540 C is needed due to the high spring back effect of the alloy at room temperature.
The temperature of each heat treatment and deformation stage must be controlled to avoid issues with the final mechanical properties. If the tube is heated above the beta transus temperature an irreconcilable loss of tensile properties occurs [ref]. However, if a high enough temperature is not reached a sufficient percentage of Ξ² phase will not form, preventing the formation of transformed Ξ² during cooling [16]. This results in a reduction in the yield strength, fracture toughness and fatigue strength of the part. The aging temperature must be controlled to ensure over aging does not occur. Whilst titanium alloys do not age in the typical fashion of aluminium and magnesium, over aging can still occur resulting in loss in hardness and fracture toughness.
4.1.2. Force Force is required during the hot deforming billet piercing and elongation stages, the warm straightening stages and the cold deforming pilgering and butting stages. In each of these stages an applied force is required to plastically deform the material, which as discussed earlier, occurs via the movement of dislocations. The influence of strain rate on the flow stress is very pronounced in titanium alloys [21]. As the strain rate is increased, the required flow stress increases dramatically as a result of strain hardening in the material. With high strain rates, work hardening occurs more rapidly due to an increase in the dislocation density [21]. This effect is a result of the reduced number of slip planes present in titanium alloys at low temperatures [21]. With limited slip planes, the movement of dislocations is greatly reduced resulting in a rapid build-up of high density dislocation regions.
During the hot deforming stages utilising an intermediate strain rate is beneficial to create a balance between the effect of the stain hardening rate and the deforming temperature on the flow stress. The strain rate must also be optimised to prevent the formation of surface defects such as cracks and ripples. These can occur when high strain rates result in a high level of friction between the tube and the working rolls.
The strain rate influences the microstructure that will be formed during subsequent heat treatments. During pilgering the grain structure is determined by the ratio at which the tubeβs diameter is reduced to rate of wall thickness reduction [15]. This ratio is optimised through the applied force to create contractile strain ratio (CSR) of approximately 1.7-1.9. This range creates a grain structure that imparts high strength, fracture toughness and fatigue strength in the material. 4.1.3. Time With respect to the heat treatment and forming stages the variable of time is directly related to temperature and strain rates respectively. During heat treatment, utilising a temperature of ~ 25-85 ΛC below the Ξ² transus, enables the shortest possible heat treatment time. During the hot deforming stages, maximising the time taken to deform the billet has a positive effect through reducing the temperatures losses but also has a negative effect in the form of potential defects, and an increased required strain rate from the presses and working rolls.
28
During quenching the transfer time between the vacuum furnace and the water bath is critical and should be kept to less than 10 seconds [ref]. This is to ensure the Ξ² phase formed during solution treatment does not decompose to Ξ± prior to aging. Additionally rapid cooling effectively traps the Al and V alloying additions in solution, increase the solid solutioning strengthening effects.
With time being linked directly to temperature, it inevitably influences the formation of alpha case and scale. If the hot deformation stages are completed over a long period of time, these surface defects will be very prominent and will reduced the fatigue strength of the tube.
4.2. Process Control A key advantage of the continuous seamless tube process is the ability to control the process variables mentioned above via simple open and closed loop feedback systems. Monitoring these variables enables instances where the variable is outside the tolerances of the set point to be detected. This is crucial in pre-emptively identifying potential defects prior to the end of manufacturing.
Prior to heating the incoming billets are inspected visually for noticeable surface defects such as chips and cracks as well as oxidation scale. This is minimise damage to the piercing and elongation mill as well as ensuring a quality tube will be produced.
Monitoring and controlling the temperature of the vacuum furnaces is critical to ensuring the correct heat treatment temperature is obtained. As discussed earlier obtain the right temperature is paramount in developing the correct microstructure and mechanical properties. A closed loop temperature control system similar to the one shown below is utilised to control the furnace temperature within +- 30 degrees of the set point.
Controller Heating Element
Temperature Sensor
Frequency Control Device
Error detector+ Set point Signal
Prior to a batch of tube hollows being formed the piercing mandrel and working rolls are checked for surface burs and cracks.
The applied strain rate used during the deformation stages is also monitored and controlled via a closed loop feedback system. Here load cells are used to detect the force being applied by the hydraulic press and the work rolls. During the cold pilgering process, the contractile strain ratio is monitored via strain radar strain gauges. This signal is feed back to the pilgering control system to notify it when to conclude the deformation process.
4.3. Quality Control A combination of inline inspection with equipment, visual inspections and mechanical testing is used to verify the quality of the products. The process control steps explained above are also used to monitor the quality of the product and are the first source of detection. If the process is seen to be operating outside of the set points, an inspection will be completed on that particular tube.
29
4.3.1. Inspection Process Visual inspections for surface defects are carried out on the tube hollows prior to the pilger mill. Dimensional checks are completed by hand to reduce costs as only rough tolerances need to be met before further reduction in the pilger mill.
After pilgering each of the tubes are inspected for defects using ultra sonic and eddy current technology. This involves passes each tube through an ultrasonic machine where the waves are used to create an image of the material structure. This is done to detect defects not visible to the human eye such as cracks and cavities. Whilst this technology is expensive it does provide a quick and effective inspection method to verify the quality of each tube. The dimensions of each tube are checked by hand using simple callipers and micrometers. This is a quick process and can there for be done manually.
4.3.2. Testing Process Destructive test work is carried out on one tube from the batch produced by the single billet. The test work involves flaring, flattening and tensile test work to calculate the rupture, compressive and tensile strength respectively. These properties indicate how the tubes will form onces joined together for the final frame.
A tube from the days production is used to test the corrosion properties via condition testing in corrosive environments. The microstructure and hardness of the tube is measured via an optical microscope and Vickers hardness testing rig.
30
5. Component Joining The final stage of manufacturing the frame is the joining of the individual tube sections. The tube sections can be joined by a number of different processes including arc welding and adhesive bonding. The common and most effective method is the TIG welding as it offers the following advantages
β’ Superior quality welds β’ Welds can be made with or without filler metal β’ Precise control of welding variables (heat) β’ Free of spatter β’ Low distortion β’ Prevents reaction of molten titanium with oxygen and nitrogen
5.1. Tungsten Inert Gas (TIG) Welding Prior to the TIG welding, the tube ends are split and shaped to enable them to fit onto the parent tube. The two tubes are then process the two tube sections are clamped together in the final positon assembly.
In TIG welding, the electrical arc is created between the workpiece and a tungsten electrode. Tungsten is used for the electrode because of its extremely high melting temperature making it ideal for service in localised heat spots. The tip of the electrode is kept sharp and precise to ensure a thigh electric arc is created. The concentrated arc between the electrode and workpiece is used to melt the workpiece [15, 18]. At the same time the technician dips a filler material, Ti-6Al-4V in the case of this bicycle frame, from a thin rod into the molten metal. As it cools a new interface is created and is known as the weld. The welding process and final product is displayed in Figure 18.
Figure 18 TIG Welding process and final weld
In the molten state, Titanium alloys readily react with atmospheric elements such as oxygen and nitrogen to form what is known as alpha case. If this extremely brittle phase is developed with in a weld the fatigue strength and structural integrity of the joint is greatly compromised. To keep these keep airborne contaminants out of the weld area during welding, the inside and outside of the joining area is bathed in an inert gas, usually argon [15]. This creates a shield around the weld preventing reactions with oxygen and nitrogen. On the inside of the frame a positive pressure purge system creates the purely inert atmosphere, whilst on the outside a modified TIG torch produces a steady stream of argon gas into the welding area.
The slightest contaminant in the weld can degrade the fatigue life of the joint. Therefor when prior to weld the frame tubes must be thoroughly cleaned. Additionally welding technicians wear cotton gloves during preparation and actually welding stages to avoid imparting oils from fingerprints into the weld.
The welding process itself is very susceptible to defects. Inappropriate weld wire movement can create vortices in the molten metal that stir up nitrogen and oxygen, allowing the elements to fuse into the weld bead compromising it in the process. Other defects include heat distortion and thermal shrinkage. These issues can result in problems with frame alignment which ultimately affects the bicycles braking and tracking ability. To minimise heat distortion the correct weld wire and bead size is needed to ensure a tight weld is created [15].
As welding involves localised melting of the melting, a change in the microstructure in the area surrounding the weld that is not melted is inevitable. In alpha-beta Ti alloys, cooling in the area surrounding the weld can result in the formation of a brittle but extremely strong alpha prime martensite phase [ref]. This changed area is known as the heat affected zone (HAZ) and unless further heat treatment is conducted it will display different properties to the rest of the frame. This difference can result in the zone becoming a stress concentration area, thus reducing the fatigue life of the joint.
Other defects that can occur as a result of welding include solidification cracking in the solidified filler material and hydrogen embrittlement. The latter can be prevented through proper shielding of the weld zone via the use of an inert gas whilst the former can be prevented with thorough cleaning of the area to be joined, avoiding high welding speeds and using the correct sized filler [22].
32
References
[1] Scott G, Buyerβs Guide: cyclo-cross bikes, Road Cycling UK, accessed 10/10/2015, http://roadcyclinguk.com/gear/cyclo-cross-bikes-buyers-guide.html/3
[2] Scott G, Charge Freezer Ti cyclo-cross frame, Road Cycling UK, accessed 10/10/2015, http://roadcyclinguk.com/gear/charge-freezer-ti-cyclo-cross-frame-nears-production.html
[3] Singletrack Forum, Which cyclo-cross frame for riding, accessed 10/10/2015, http://singletrackworld.com/forum/topic/which-cyclocross-bike-or-frame-for-racing
[4] Princeton College, Princeton Bike Design Issues, Accessed 08/10/2015, https://www.princeton.edu/~humcomp/bikes/design/desi_70.htm
[5] Silva V D, Mechanics and Strength of Materials, Springer, New York, 2006
[6] Dwyer F, Shaw A, Material and Design Optimization for an Aluminium Bike Frame, Worcester Polytechnic Institution, 2012
[7] Brown S, Frame Materials for the Touring cyclist, Accessed 09/10/2015, http://www.sheldonbrown.com/frame-materials.html
[8] Performance Bicycle, Basic Guide to Bike Frames, accessed 11/10/2015, http://learn.performancebike.com/bikes/advice/buyers-guides/bikes-and-frames/basicguide-to-bike-frame-materials
[9] How Products Are Made, Bicycle, Volume 2, accessed 11/10/2015, http://www.madehow.com/Volume-2/Bicycle.html
[10] Ibis Cycles, Metallurgy for Cyclists, accessed 12/10/2015, http://www.ibiscycles.com/support/technical_articles/metallurgy_for_cyclists/the_basics/ [11] Groover M P, Fundamentals of Modern Manufacturing 4th Edition, John Wiley & Sons Inc,
[12] Bralla J G, Handbook of Product Design for Manufacturing, 1986, McGraw Hill, New York, USA
[13] Ashby M F, Materials Selection in Mechanical Design, Butterworth-Heinemann, Oxford, 1999
[14] Polmear J, Light Alloys, Butterworth-Heinemann, Oxford, 2006
[15] seven cycles, Sevenβs Approach to Frame Building, accessed 11/10/2015, http://www.cyclefit.co.uk/uploads/Seven_Titanium.pdf
[16] Donachie M D, Heat Treating Titanium and its Alloys, Titanium: A Technical Guide, ASM International, Ohio, 2000
[17] Titanium Exposed, Titanium Extrusion, accessed 12/10/2015, http://www.titaniumexposed.com/titanium-extrusion.html
[18] Froes F H, Titanium: Physical Metallurgy, Processing, and Applications, Worcester Polytechnic Institution, 2012
[19] Hashmi M S J, Comprehensive Materials Processing, Newnes, 2014
[20] Ubhi H S, Houghton A, Saithala J, An EBSD study of Texture Variation along Pilger Reduced Titanium Alloy Tubes, Oxford Instruments,
[21] Deiter G E, Kuhn H A, Semiatin S L, Handbook of Workability and Process Design, ASM International, 2003
[22] TWI, Defects-solidification cracking, accessed 15/10/2015, http://www.twi-global.com/technical-knowledge/job-knowledge/defects-solidificationcracking-044/
34