3-2.5 versus Other Materials

Steel

Although the ultimate tensile strength of many premium steels is greater than 3-2.5 titanium, this raw strength is meaningless in the final bicycle frame because:

1. The strength advantage is lost in welding.
2. Steel's strength-to-weight ratio is lower than that of titanium, both before and after welding.

When comparing materials, strength after welding, or heat-affected strength, must be considered first, because the highest stresses in a frame are at the joints or heat-affected zones. For example, Columbus SL steel tubing has a cold-worked (as received) ultimate tensile strength of roughly 135 ksi, making it equal to Merlin 3-2.5. Ignoring for a moment that Merlin's strength-to-weight ratio is almost double that of the Columbus SL, we find that SL's yield strength drops to 70-78 ksi after welding. Merlin 3-2.5 has a post-weld yield of 97-100 ksi. In addition, for a given weight 3-2.5 titanium has roughly twice the post-weld fatigue strength of 4130 chrome-moly steel.

External and internal reinforcements, such as gussets, butts and lugs, can improve steel's fatigue strength somewhat. Internal butts move the weakest points away from the areas of highest stress. In some cases, however, it is not possible with current manufacturing equipment to create a butt of optimum thickness. The maximum differential between the butted and unbutted sections of a production premium steel tube is about 40%; any further improvement must be achieved in some other way-with gussets, lugs, or some variant of these.

An optimally butted steel tube will outperform a gusseted or lugged tube because:

1. A gusset or lug does not reduce the heat-affected zone (HAZ) at the sides and end of the reinforcement. An ideally butted tube provides equal strength and equal or lower weight with no HAZ.
2. Gussets and lugs create stress raisers at their endpoints, with a further reduction in fatigue life due to the HAZ. An ideal butt with a properly designed taper eliminates the stress raisers and also saves weight.

Whether gussets and butts are employed or not, there is still a wide gap between the fatigue strength-to-weight ratio of 4130 steel and 3-2.5 titanium. Claims that it is possible to create a steel frame of comparable weight and strength as a titanium equivalent are unsupportable, as proven by raw objective data, and by the fact that no such frames exist.

Aluminum

Unlike titanium, aluminum's fatigue strength declines continuously with increasing cycles. Therefore, aluminum designs must include a greater design safety factor, which inevitably increases weight and bulk.

A related issue is the failure mode of aluminum, which is catastrophic, rather than gradual. Again, the design safety factor must be increased to compensate.

Aluminum is a good material for low-stress components that see little to no fatigue cycling.

MMC (Metal Matrix Composites)

There are many available types of MMCs, but only one, a particulate-type from Specialized/Duralcan, is presently being used in bicycle frames.

Particulate-type MMCs are the least-expensive form in current production. The Duralcan MMC is an aluminum oxide particulate matrix in an aluminum medium. Other MMCs under development for bicycle use are also particulate types. One employs silicon carbide, the other boron carbide, both in an aluminum base.

MMCs vary in the base metal from aluminum to titanium to copper, and in matrix additives as noted above. The formats of the additives also vary, from particulates, whiskers and wires to continuous and discontinuous fibers. Each factor plays a large part in the strength and other mechanical properties of MMCs. One thing common so far to all particulate and whisker MMCs is a loss of ductility and fracture toughness, which has had a negative effect on potential fatigue life.

Duralcan's 6061-T6 15% particulate MMC has the following advantages over pure 6061-T6 aluminum:

Tensile modulus is increased 30% to 12.7 ksi. The higher modulus helps offset the material's low fatigue life, since a stiffer frame has a lower stress cycle.

Yield is increased by 15%, from 40 to 46 ksi.

Disadvantages of the Duralcan MMC include:

Elongation hovers at a meager 5.4%, potentially decreasing fatigue life. (Theoretically, if the frame were designed for no flexure whatsoever, elongation would not affect fatigue life, since the joints would not move. In practice, however, this seems unlikely.) Elongation drops another 50% or more for other MMCs. The lower the number, the less ductile the material. 6061-T6 aluminum has 14-17% elongation after welding and heat treatment. High-quality bicycle steel is 10% before welding, 20-25% after welding. Titanium's elongation is 10-19% before and 15-30% after.

Stress vs. Number of cycles (S-N) fatigue curves remain almost identical to off-the-shelf 6061-T6: 17 ksi at 107 cycles for Duralcan MMC vs. 16 ksi for 6061-T6. Therefore, the fatigue strength-to-weight ratio is almost identical to standard 6061-T6. Note that this fatigue strength is hypothetical because, like monolithic aluminum, MMCs do not have true fatigue endurance. Instead, they must be designed with a much more conservative safety factor.

Fatigue strength is the most important consideration in frame design, regardless of which frame material is under consideration. Most frames fail through fatigue, not from one-time overloading, as in a crash. Ultimate strength is of secondary importance, because a high UTS alone does not and cannot make a durable frame.

The most obvious theoretical benefit of any MMC is the potential to create a stiffer material, as in an engine block where rigidity can reduce noise and vibration. This, however, is not necessarily desirable in a bicycle frame. Ride quality is an important consideration that must be incorporated, even if the fatigue issues are satisfactorily resolved.

Welding is also a complication. Most MMCs lose strength after welding, and some of that strength remains unrecovered after heat treatment. In the area closest to the weld (as well as in the weld itself), the particulates become dispersed, which can cause anomalies and strength problems. Heat treatment cannot restore these particulates to their pre-welded state because the metal does not liquefy during heat treatment.

Finally, it should be noted that a bonded MMC frame can never match the weight of a welded MMC frame. Thus, it is doubly unfortunate that many MMCs have serious mechanical degradation after welding.

Titanium Metal Matrix Composites

There are very few titanium-based MMCs in current production, with only two basic types of matrices. One, intermetallic-matrix composite (IMC), uses continuous fiber. The other is formed from titanium carbide particulates. Both have been developed primarily for high-temperature applications, as in engine components and skins for military aircraft.

IMCs are formed from a series of titanium-aluminide foils consolidated with boron-coated silicon carbide continuous fibers. With a starting price of $2000-3000 per pound, it is unlikely they will soon find applications in the bicycle field. Interestingly, raw ingots of titanium cost only $10-12 per pound, so the processing costs to create IMCs are obviously formidable.

Titanium carbide MMCs present similar cost issues. They also suffer from a severe loss of ductility which arises from the induction of carbon into titanium.

Titanium-aluminides are another newly publicized group of aerospace alloys. Strictly speaking, these are not MMCs, but they do boast very high strength and good resistance to loss of mechanical properties at high temperatures. However, they suffer from abysmal ductility at room temperature and exorbitant cost. The ductility issue may soon be resolved; cost, however is unlikely to drop within the foreseeable future.

Beryllium

Beryllium is a light, stiff, and expensive metal that has received recent attention as a potential frame material. Merlin began cooperative work with a beryllium tube manufacturer two years ago, but our preliminary investigation revealed that the stiffness-to-weight ratio of beryllium is extremely high-so high that it would be difficult to build a frame with adequate flex for good ride characteristics. Furthermore, beryllium's cost is so prohibitive that the financial wherewithal necessary to develop a frame is beyond the resources of the bicycle industry.

Even those alloys that incorporate beryllium as their major element are so expensive that it is doubtful any of them will ever find their way into the frame tubing market. In addition, beryllium is toxic, although this can be managed with proper manufacturing procedures.

Carbon Fiber

Carbon fiber is a blanket term for a wide variety of carbon-impregnated polyesters, graphite fibers, and polymerized carbon fibers that are used within a matrix of adhesive to create a clothlike structural material.

Within the family of fibers considered appropriate for bicycle frame use, the raw fibers' stiffness-to-weight ratio is roughly 3.5 times higher than 3-2.5 titanium. The ultimate tensile strength is roughly 70% higher.

However, these figures apply only to the raw fiber strand, before it is impregnated by and retained within an epoxy resin matrix. The epoxy adhesive's structural properties are significantly lower. Moreover, epoxy normally occupies 50% or more of the cross-sectional area of a sheet of carbon fiber cloth. This ratio of resin to carbon must be maintained to hold the fibers together; a lower epoxy content reduces the fiber weave's layer-to-layer shear strength. A 50% volume of adhesive reduces the finished product's strength-to-weight ratio by a factor of two.

In addition, carbon fiber is anisotropic, which means that it displays directional properties. For example, a fiber with a modulus of 20,000 ksi when measured longitudinally will have, at best, a transverse modulus of 4,000 ksi. Similarly, the ultimate tensile strength may measure 220 ksi longitudinal, but will be, at best, 10 ksi transverse.

This anisotropic property can be exploited beneficially in some structures, such as leaf springs. However, bicycle tubes must be able to carry stress loads in many planes at once-in tension, compression, fully reversed bending and clockwise and counterclockwise torsion. Thus, it is virtually impossible to utilize anisotropy to any significant extent in a frame.

In addition to the modest structural properties displayed by the epoxy resin, carbon fiber has extremely low ductility and poor abrasion resistance. Historically, low ductility in those bicycle frames that do not use separate lugs has led to joint failure and stress cracking. Abrasion is a particularly thorny problem since composites are notch-sensitive, such that even minute inconsistencies in the material can develop into large cracks, eventually leading to failure.

Abrasion problems can be reduced at the cost of added weight by a protective skin or veil of fiberglass or, at higher cost and somewhat greater strength, Kevlar fiber, but the abrasion resistance of these and similar polyester and aramid fibers is also low. Abrasion and impact damage can be repaired with epoxy-based fillers and additional cloth. However, since the integrity of the structure is dependent upon continuous fibers in tension, the strength of the repaired area will be lower than the original material, and the weight of the repair will be higher.

Carbon-wrapped Titanium and Aluminum

Titanium or aluminum tubing wrapped with a bonded layer of carbon fiber composite has been proposed as a method to achieve a synergistic improvement of material properties. (In fact, carbon-wrapped aluminum tubing was produced by Easton for Raleigh for two years, before the withdrawal of that frame from the market.) The main objectives of this approach are:

1. To improve the performance of a low-strength tube. Aluminum's low strength-to-stiffness ratio, for example, can be boosted appreciably with a layer of high-modulus composite fiber.
2. To protect the abrasion-sensitive carbon within a metal exoskeleton.

These approaches have a number of drawbacks:

1. External carbon wraps do not solve the problem of abrasion damage to the composite.
2. Internal carbon wraps do not necessarily protect the composite from impact failure either. To create a frame of reasonable weight, the titanium or aluminum tube must be very thin, and consequently not resistant to denting. Since titanium is very ductile, it can spring back from minor impact with no appreciable damage. However, the internal wrap will suffer local cracking, which can spread into a serious fault.

In addition, delamination of the composite from the tube surface is a serious long-term problem. It has at least three sources:

1. Delamination can occur from impact. Once the composite has cracked, it will continue to fail along the fiber orientation. The fissure created by the initial fault becomes a point for peeling or cohesion.
2. Delamination can occur at the ends of the supporting tube due to applied bending and torsion during use. Adhesives are weakest in peeling and cleaving.
3. Delamination can occur from stress. When used in a wrap, the adhesive must perform two duties, first as the bonding agent between the fibers, and second as the glue between the composite and the tube. Ideally, two different adhesives and primers would be specified, but this is not always possible.

Carbon-wrapped tube frames also suffer from a weight disadvantage, since these tubes cannot be welded once the composite has been applied, and so must be bonded in a lugged frame.

Honeycomb-Reinforced Titanium

Honeycomb-reinforced titanium tubing is conceptually similar to internally wrapped composite tubing, with the primary objective being increased stiffness. The only frame that currently employs this construction uses a lightweight fiberglass honeycomb bonded to a carbon fiber skin, which in turn is bonded to the inside wall of a thin titanium tube. The frame is lugged.

In the current design, the honeycomb lends anisotropic reinforcement properties to the tube. Unfortunately, it is not possible to create layers of directional honeycomb, as can be achieved with carbon fiber. Thus, the honeycomb is inevitably unidirectional, but lies within a structure that demands more isotropic properties.

Since the frame must be lugged for assembly, frame weight is not ideal; a current 54-cm example weighs 3.0 pounds, with the honeycomb and carbon representing 0.75 pounds of this total. A 54-cm Merlin Extralight, with double-butted tubing and similar rigidity, weighs 2.6 pounds.