36 replies to this topic

[KinetiK]

What are the usual materials used for the torsion bars? I've heard that some teams use titanium or a titanium/maraging steel mix. Has anyone here heard of this material: 5S99G? I can't find a reference to this material in any ASTM charts, however my ASTM lists are pretty old and this is apparently a newer material.

[desmo]

I am conducting further inquiries but at present it appears likely that steel torsion bar springs are currently employed. In issue 29 RaceTech reported that the McLaren MP4/15 used steel as the material for it's springs. Prior to the changeover to the current torsion bar-type springs Ti alloy coils were the standard. I am currently trying to find out why the teams have apparently reverted back to steel with the changeover to torsion bars. I have found reference to Ti-alloy torsion bar-type springs for aerospace applications, so it appears that there is nothing intrinsically unsuitable about Ti-alloy for this type of spring in general. There appears to be some problem with it's use in this application however.

[KinetiK]

One of the teams used maraging steel/titanium combination but the results "were not predictable". That material I mentioned, 5S99G, in used by the same team. I'm a beginning materials buff and much to my chagrin, I was unable to find ANY information about it on the internet. There is quite a bit of information regarding maraging steel but most of that is based on uses for golf clubs. I'm eagerly awaiting any information you might have Desmo!

[Top Fuel F1]

Originally posted by desmo
it appears that there is nothing intrinsically unsuitable about Ti-alloy for this type of spring in general. There appears to be some problem with it's use in this application however.

Other than Ti's obvious weight advantage over steel it's main advantage is it's ability in valve spings not to Take Set any where as fast or to the degree that steel does. However there could be some problem concerning the forces uniquely involved with torsion. Ti's obvious disadvantage would be it's cost. I imagine especially so in a set of torsion rods having so much more material than than a set of valve springs. Possibly the forces are such on the torsion rods that any Set they may take during a given race weekend is tolerable and therefore they just change them out at a time dictated by the teams experience. The weight may be a moot point as long as the torsion rods are not causing the car's minimum weight to be exceeded and the weight is distributed to the team's liking.

Rgds;

[desmo]

One thing for sure is that cost is irrelevant as far as springs in F1, and another is that weight is NEVER moot. I'd imagine the teams would be thrilled to pay 100USD/g for a useful weight reduction and no engineering material-even Be- costs near that much. If the teams are using steel springs as it appears they likely are, then I'm sure there are sound engineering reasons behind it and that is all there is to it.

[Halfwitt]

5S99G is commonly referred to as S99. The 5 is the fifth revision of the standard, and G refers to 'bright bar' i.e. cold drawn or peeled etc.

It is a fairly standard steel: carbon, nickel, moly, chrome are the main ingredients. It is a cleaner version of the old EN25 specification, and has a 'fairly low' tensile strength of 1230 - 1420 MPa.

I don't know what diameter they use for torsion bars, but standard spring steels are much harder (and thus stronger).

[Marco94]

Steel has a higher elasticity modulus than Titanium. So for a given stiffness, a steel torsion bar should be smaller. Not sure if it would be lighter. BUt it's smaller dimensions could lead to better packaging properties.

Marco.

[david_martin]

The shear modulus of steel is around 80 GPa and the density around 7850 kg/m^3. For titanium the figures are around 44 GPa and 4500 kg/m^3, respectively. In the simplest case (neglecting fatigue and yield strength considerations), whether a Ti or steel torsion bar of a specified torsional stiffness is lighter can be determined from the following:

 G(steel)	   Rho(steel)

----------- >  -------------

G(titanium)	Rho(titanium)

ie. if the ratio of the shear moduli is larger than the ratio of the densities the torsion bar will be lighter. For this case we get 1.81 > 1.75, hence a steel torsion bar should be lighter than an equivalent titanium one.

[Marco94]

There you go. I didn't have the numbers around, so thank you for pointing this out.

[KinetiK]

Halfwitt

Thanks for the information about the 5S99G steel.

[Yelnats]

The weight of Torsion bars (or any other material) in an F1 car becomes more important as it moves upward in the chassis and away from the center of mass. If the Torsion bar placement satisfies these criteria (low and near the center of mass), it's weight becomse far less important in terms of ballast compensation. The torsion bars may ultimately be very near the ballast anyway so why lighten them at great expense and design complication when one is only going to place an equivilant mass of ballast nearby?

In the case of a transmission/engine which is placed well back in the car and relativly high up in the chassis a reduction in mass is always going to improve the balance of the car by allowing more flexibilty in ballast placement to improve the Center of Gravity and Mass. This may not always be true of Torsion Bars weight savings.

[Top Fuel F1]

Originally posted by Yelnats
The weight of Torsion bars (or any other material) in an F1 car becomes more important as it moves upward in the chassis and away from the center of mass. If the Torsion bar placement satisfies these criteria (low and near the center of mass), it's weight becomse far less important in terms of ballast compensation. The torsion bars may ultimately be very near the ballast anyway so why lighten them at great expense and design complication when one is only going to place an equivilant mass of ballast nearby?

Well said Yelnats. This is what I had in mind on my previous Post. That is the weight savings may be moot if and only if certain circumstances exist.

[Top Fuel F1]

Originally posted by Top Fuel F1
However there could be some problem concerning the forces uniquely involved with torsion.

So as it has been well pointed out by several people Posting, it's the difference in the Shear Modulus that that would lead you to believe the F1 torsion rods are a steel alloy. It seems what makes Ti a more desirable material than steel for valve springs (CART, Top Fuel, and maybe IRL) as peported by RCS:
What makes titanium so much better than steel? Quite simply, titanium is able to handle the stress of repeated compression without "bending" (Taking Set). The difference is in the "Yield Point" and "Fracture Point" of the two materials. The chrome silicon alloy used in normal clutch springs has a yield point (at which it bends) and a fracture point (where it breaks) that are rather far apart. Titanium on the other hand has a yield point and fracture point that are much closer together. The other attraction to titanium is the light weight. Compared to steel, a titanium spring typically requires ½ the volume and 1/3 the weight. As well the wire dia. of the Ti spring wire can be less than the dia. of a steel one, enabling a higher lift capability.

Rgds;

[Cory Padfield]

David Martin's post shows an equation for "the simplest case (neglecting fatigue and yield strength considerations)". Why would one neglect yield strength considerations? A spring (torsion, axial, coiled, whatever) will be a useful one until its elastic limit is reached. The actual performance index P for a minimum mass spring is:

P = (sigma)^2
--------------
E * rho

where sigma is the yield strength, E is the Young's modulus, and rho is the mass density.

Using the following realistic material properties:

Property | Steel | Titanium
-------- | ----- | --------

sigma | 1200*10^6 Pa | 1200*10^6 Pa
E | 207*10^9 Pa | 110*10^9 Pa
rho | 7.9*10^3 kg m^-3 | 4.5*10^3 kg m^-3

P | 880.6 m^2 s^-2 | 2909 m^2 s^-2

Spring mass is inversely proportional to P. Therefore, a titanium spring can be significantly less massive for a constant input force.

I cannot authoritatively comment on why F1 teams do not use titanium alloy torsion bars, but the reason(s) could include:

ignorance, lower fatigue strength, volume constraints (a less massive titanium spring will require larger volume to resist the same input force)

Best regards,

Cory Padfield

[desmo]

So far no one has shed the slightest ray of light on why Ti alloy would be better than steel for coil-type springs- which I've heard described as coiled torsion bars- and at the same time be a worse material for torsion bars. Is there some fundamental difference in the desired properties of materials for the two types of springs? And when I suggested (in the Naked Arrows thread) that a Ti-alloy spring would lose much of it's weight advantage vs steel as it would require a greater volume of material to compensate for it's lower elastic modulus, I was "corrected" yet Cory makes essentially the same point: "(a less massive titanium spring will require larger volume to resist the same input force)". What am I missing here?

[imaginesix]

Good Q desmo, and to the point.

I also always thought of coil springs merely as torsion bars 'coiled up'. The bar DOES twist when the spring is compressed, after all.
But this can't be the only resisting force, as I would think that a coil spring wire is much thinner and longer than a torsion bar of equivalent force, if you were to uncoil it.
Perhaps there is some kind of 'multiplication' to the torsion of a coil spring or in it's connection to the suspension. Perhaps this also explains the different material requirements for the two spring mediums.
Of course I have no technical explanation, so I too would like to hear from the pros out there what that explanation might be.

[Cory Padfield]

A spring (regardless of shape) with minimum mass is one with the maximum performance index P = (sigma)^2/(E * rho).

Using typical materials properties, titanium alloy springs will be less massive than steel alloy springs, which provides an obvious reason for using titanium alloys.

Now, why would F1 teams not take advantage of less massive springs? My best guess is that titanium springs may require larger volume, which is at a premium in the vehicle. Shaft torsion obeys the equation

tau = G * theta * r / l

where tau is shear stress, G is shear modulus, theta is angular displacement, r is shaft radius, and l is shaft length. Now, which of the symbols sigma/tau, E/G, rho, theta, r, and l are fixed and which are variable? When you answer this, you can make the optimum material selection for your vehicle.

Best regards,

Cory Padfield

[desmo]

From the formula posted by David Martin in the thread, it would appear that given typical physical properties for Ti alloys and steel, the steel spring should be slightly lighter than a Ti spring of equivilent spring rate. And yet we have this from the SAE website:


Because of titanium's low shear modulus,
there are not as many turns required for
titanium springs as steel springs. Even if the
two materials have the same density, the
titanium spring would be lighter because it
does not use as much material. Since the
density of titanium is about half that of steel,
titanium can perform the same task as steel
springs in most applications while weighing
60-70% less.

So it appears that titanium's low shear modulus allows it to be made into coil springs massively lighter than equivilent rate steel springs, yet this advantage appears to largely evaporate when the spring type is a torsion bar. Either the SAE site is giving out misinformation on the subject, or the formula posted by David Martin for calculating a spring's mass for a given rate is failing to describe the case of a coil-type spring. Am I the only one that sees a contradiction here?

[Cory Padfield]

The formula posted by David Martin is not accurate.

The SAE website is correct.

It is understandable that there is confusion and perceived contradictions because materials and shape selection can be complex.

Perhaps my previous posts are unclear, so let me try again. Assuming typical properties from current materials, titanium alloy springs will be less massive than steel alloy springs regardless of shape. To take full advantage of a material's properties, the geometry must be allowed to change. However, in many applications, the geometry is fixed (or constrained with a narrow range) rather than variable.

Engine valves, for example, would seem to have fixed displacement (to meet required valve lift), fixed coil diameter (to avoid touching head walls, other springs, etc.), but have variable elastic modulus, number of turns, and wire diameter.

For suspension torsion bars, it would seem that theta is fixed (minimize wheel travel/pushrod displacement), but modulus is variable. What about shaft radius and length? Are they variable, or are they fixed? If they are fixed, this may mitigate some (or all?) of titanium's advantage.

If there is interest, perhaps we could do some sample calculations if reasonable estimates for geometry and displacements are available. I saw in another post that "suspension travel" numbers are available. I am sure valve lift numbers are also available. If we could get dimensional constraints, we can design our own springs and compare steel to titanium. Here is a list of necessary values:

Torsion bar: theta (angular displacement, which could be calculated from tangential linear displacement is we know how much the pushrod changes), bar radius constraint, and bar length constraint

Valve spring: axial displacement/valve lift, wire diameter constraint, coil diameter constraint, and number of turns constraint

I believe that is all. I am interested to see where this goes...

Cory

[desmo]

After scrutinizing Piola's drawings of the F399 torsion bar rear suspension from his Technical Analysis '99 book, I made the following estimates of the length, diameter and angular displacement of those springs.

Estimated length: 320mm

Estimated diameter: 13mm

Estimated angular displacement at full travel: 20-25 degrees

The current car's rear suspension looks similar to me. These springs are tiny and cannot weigh much. Frankly it doesn't look like there's a whole lot of weight to lose in the springs. If these things weigh much over a kg, I'd be surprised. It's amazing to me they can get the required rate out of such a small quantity of metal. The springs are splined on both ends and have what appears to be a bearing surface at mid-length. I wonder how much their length contracts by at full twist.

[Cory Padfield]

So, we have the following information:

steel torsion bar

G = 76 * 10^9 Pa
theta = 0.436 radians
r = 6.5 * 10^-3 m
l = 320 * 10^-3 m
tau = 673 * 10^6 Pa
mass = 335.5 * 10^-3 kg


If we just said "hey, let's make the exact same torsion bar from a block of titanium instead of a block of steel", the we would have:

titanium torsion bar

G = 42 * 10^9 Pa
theta = 0.436 radians
r = 6.5 * 10^-3 m
l = 320 * 10^-3 m
tau = 372 * 10^6 Pa
mass = 191.1 * 10^-3 kg
mass decrease = 49%


But, titanium has essentially the same strength as steel, so we are under-utilizing the titanium bar. If we load it to 673 MPa, then we have:

2nd titanium torsion bar

G = 42 * 10^9 Pa
theta = 0.436 radians
r = 6.5 * 10^-3 m
l = 177 * 10^-3 m
tau = 673 * 10^6 Pa
mass = 105.7 * 10^-3 kg
mass decrease = 68%


This verifies the SAE website information. It also verifies the performance index criterion (P_Ti = 3 * P_St & m_Ti = 1/3 * m_St, inverse relationship). Now that we have done the analysis, I am unsure why F1 teams don't use titanium alloys. Perhaps it is a material property issue (fatigue strength, ease of fabricability, etc.), or ????

Cory

[Cory Padfield]

Here are a few more thoughts...

My previous analyses assumed properties for medium-alloyed steels and alpha + beta titanium alloys, however maraging steels and newer beta phase titanium alloys are certainly available to F1 teams. Here is a new analysis:

Property | Steel | Titanium
-------- | ----- | --------

sigma | 2300*10^6 Pa | 1250*10^6 Pa
E | 200*10^9 Pa | 100*10^9 Pa
rho | 8.1*10^3 kg m^-3 | 4.76*10^3 kg m^-3

P | 3265 m^2 s^-2 | 3283 m^2 s^-2


So, the gap is much smaller. Other factors (machinability, fracture toughness, fatigue strength) would play a major role. Personally, I never think of maraging steels at the same time as "regular" steels because of their significantly different properties (including cost). Maraging steels, along with titanium alloys, are the best performing metals for mass-critical parts.

Cory

[imaginesix]

Wow. Information overload!

...give me more, I want MORE!!!

[RDV]

...Ti is more notch sensitive than steel, it galls very easily,there is no weigth advantage over steel, (we are not going to quibble over grams ) Ti is a great material on forgings , but quite fiddly to work with.

Uprights give you a reasonable weight saving but on torsion bars steel is still the best cost /weight/efficiency ratio. Not a mass critical item.

[RDV]

....forgot to mention that on engines the Ti springs have better damping for same forces, usefull to reduce surging at hi revs, but not really relevant in F1 as pneumatic, but back to torsion bars what is important is specific stiffness, for most metals this is @ 3.7 ,moving around in the second decimal place area , steel -Ti,ally are all practicably interchangeable,packaging, fatigue life, machining and volume are all factors that determine what material to use. Ti useful on hubs as thermal conductivity is low , keeps bearings cooler , but again function specific.

For truly great savings in weight go to a carbon arb. (specific stiffness @17 to 20) as in the 963 porsches and the GT1 .

[Cory Padfield]

RDV,

Titanium valve springs are used because they have less mass for a given spring displacement - it has nothing to do with damping.

Torsion bars are not dependent upon a materials specific stiffness (elastic modulus divided by density). Their dependence is specific strength (fatigue strength squared divided by elastic modulus).

What is an arb.?

Cory

[desmo]

ARB= Anti Roll Bar...another spring.

[Top Fuel F1]

Originally posted by Cory Padfield
RDV,

Titanium valve springs are used because they have less mass for a given spring displacement - it has nothing to do with damping.

Cory

Cory:

In the NHRA Top Fuel ranks the general engine guys say they are using Ti valve springs because the steel ones "Take A Set" so quickly. Possibly as soon as the first burn-out. As a materials engineer, I would appreciate it if you could elaborate on that.

Rgds;

[RDV]

Torsion bars are not dependent upon a materials specific stiffness (elastic modulus divided by density

...beg to differ, if Im trying to get the stiffest gizmo for the lightest weight , this is how I factor it out, else why use carbon wishbones ? as I mentioned before choice of material depends a lot on specific application and manufacturing method.

When we went to ally honeycomb composites the fact that panels were much stiffer for a given weight was counterbalanced by the fact that as outer skins were much thinner meant riveting; the then current paradigm on chassi manufacturing,was not very practical, shifting to bonding brought rolled tube ally composites to the fore, material and manufacturing method combined to produce a well engineered product

material use is application related, and is very cyclic, we had a rash of Ti everything in the 70`s , even on items that had no concievable engineering gain by using it, it was the in material , just as carbon is now. MMC and beryllium likewise. I have seen gear lever knobs in Ti.... justify that one....

Carbon is very usefull , but is used sometimes on items that could be made in some other material just as well , but sound sexier if in latest unobtanium.

Even if you have a good materials wonk on the team, F1 is just as much of a fashion victim as any teenager, hard to justify your claims to be on the cutting edge of technology when metals used in car are mainly steel, why ,washing machines and rails are made of it....

This years word is.......ceramics.

Know also of F1 team that had a great deal of difficulty making wishbones in carbon last, so they merrilly ran steel wbones with a carbon-look sticker over it for nearly half a season....

re spring surge damping with ti , have never delved into it myself , just passing on what a engine wonk told me upon my enquiring.....am mainly aero and vehicle dynamics man.... engines are just lumps you put in cars to push them along......

but as far as use in suspension goes , using steel springs versus TI ones produce no gain in weight again through specific stiffness ,

Ti springs used to be more common because until racing car springs started being sourced and manufactured with high quality steel using ground wire and shaped close to the elastic limit and Wahls limit they used to be made in set steps of wire gauge, and had more or less turns to give required rate.

Ti ones which cost one order of magnitude more were much better engineered, but with a volume penalty, were lighter as designed to suit.

When springs started to be sourced from aerospace manufacturers they were ground to specified size and done on limit of material ; they were equivalent to Ti (within a dozen grams per car set..)

Sensible engineers specify any weight gaining exercise in $/Kg or suitable local currency and mass unit Chapmans dictum of a (pound sterling)/( pound mass) still holds, even if the numbers have changed considerably upwards.. so QED, on current torsion bar size and application verdict goes to steel

[Jezztor]

I have to agree with RDV, seen it in ISRS series before too, and recently read an article about this.

Know also of F1 team that had a great deal of difficulty making wishbones in carbon last, so they merrilly ran steel wbones with a carbon-look sticker over it for nearly half a season....

Yeah, BAR only made the jump to full-carbon this season.

I would agree that this years fashion accessory is ceramic. I recently read an article about a firm who is trying to incorporate some sort of ceramic strand into composites, aiming at rear suspension heat resistance for F1 cars (exhaust heat protection). Ceramics are relatively cheap, in comparison to the likes of gold, which Ferrari were / are (?) presumably using, as well as other teams. Exhaust heat, to my knowledge, retains about 600 degrees Celcius upon exiting the exhaust. Anyone got any more info on this company?

CF has been the 'in-thing', as you rightly said, RDV, since the late 70's / early 80's, and then I think it was Chapman who went and built the first full-CF F1 car in mid 80's. In those days CF was a bit of an unknown, no one really knew how it worked; no one really knew how to use it properly, and had to experiment...but the use of CF even in those days was very image-orientated.

Jezz

[RDV]

.....gold is cheap, it just has this image....... it is extensively used in electronic components as does sterling (tongue in cheek....) work with contacts, resists spark erodind and oxidation, good temperature capabilities, and when used on carbon suspension members is best material/weight /efficiency compromise.... cost not an important factor, several other engineering materials can be just as expensive to use. diamonds also have their image , but very efficient in some types of grinding wheels!!

ceramics have this cheap image as we all seem to equate ceramics = material flower pots are made of, some ceramics can be very expensive...

[Top Fuel F1]

Originally posted by Top Fuel F1


Cory:

In the NHRA Top Fuel ranks the general engine guys say they are using Ti valve springs because the steel ones "Take A Set" so quickly. Possibly as soon as the first burn-out. As a materials engineer, I would appreciate it if you could elaborate on that.

Rgds;

Re:

http://www.coilsprin.../dragvalve.html

[RDV]

....doesnt make much sense , sounds more like a sales pitch, will try to find info from Schmidthelm , who used to make most of the springs for Cosworth , and mainly the Ti ones

[Toby Padfield]

I'm going to use the analytical concept "Performance Index" as described by M. F. Ashby in his book Materials Selection in Mechanical Design to describe spring performance. This analysis does not address lightweight designs for suspension links or anything else where the performance index is not the same.

A spring is a device for storing elastic energy. The elastic energy stored per unit volume when stressed to a nominal stress Sigma is (1/2 * Sigma^2)/E. If failure occurs when Sigma exceeds a failure stress Sigmaf, then the maximum energy density is given as W = (1/2 * Sigmaf^2)/E.

This equation describes an axial spring. Since solid torsion bars have a lot of material at the neutral axis that is not stressed highly, the energy density becomes W = (1/3 * Sigmaf^2)/E. Note: Ashby continues the derivation using E, the Young's modulus, instead of G, the shear modulus. For isotropic materials, G is related to E by the relationship G = E/[2*(1+nu)] where nu is Poisson's ratio. Based on this, one can see that to minimize the volume of a spring, the best material would maximize the quantity Sigmaf^2/E. In order to minimize the mass, density now enters the equation: the performance index to be maximized is now Sigmaf^2/(rho * E).

The failure stress to be used for any dynamic spring, such as suspension springs/torsion bars for automobiles and racing cars, is a fatigue strength at some number of cycles. Obtaining this type of data in the exact same format (stress ratio, stress amplitude, etc.) is fairly difficult. As a proxy, I'll use ultimate tensile strength (UTS) numbers:

Oil-tempered, Chrome-Silicon steel, Valve Quality (ASTM A 877): UTS = 2070-2240 MPa for a wire diameter of 1 mm.

Beta-C titanium alloy (Ti-3Al-8V-6Cr-4Mo-4Zr), solution treated, cold drawn 90% to ~ 1mm diameter: UTS ~ 1500 MPa. Subsequent aging increases UTS to 2000 MPa. This data is from SAE Techical Paper 890470 (Honda, Chuo Spring, & Suzuki Metal Ind. Co.) Beta C is usually supplied to a range of strength specified in the aerospace standard SAE AMS 4957, but I don't have access to this particular standard, so I used the Honda data.

From this data, it is not surprising to see why titanium can substantially reduce the mass of springs-- the strength can be nearly the same, but the density is only 4.82 g/cc instead of 7.83 g/cc, or 52% that of steel.

Things to consider:
1. Steel will retain a higher fraction of its tensile strength under fatigue conditions that feature mean stress effects. The proportion isn't high enough to overcome the large difference in density, though.

2. The Performance Index method accounts for optimization of spring parameters like coil diameter, number of turns, etc. in order to minimize mass. However, placing constraints on block height (length when fully compressed, also called solid height), free height (length when unloaded), etc. can greatly influence stresses on the wire, which is why changing from steel to titanium and vice versa can result in mass savings. The article that Top Fuel posted earlier is a good example of this. Another example is from a Ford study (summary appears in 1997 issue of JOM, p. 40):

steel titanium
Rate (N/mm) 61.25 61.25
Free length (mm) 426 426
inside coil dia (mm) 102 102
max load (N) 13,642 13,642
max stress (MPa) 1,031 872
wire dia (mm) 16.95 18.00
active coils 7.90 4.97
total coils 9.09 6.16
mass (kg) 5.92 2.86

So, for a given free length, the steel can't get out of it's own way-- the stress increases too much, because it is too stiff. This is why "setting" can be such a problem with springs: stresses build rapidly as compressed length approaches solid height, which isn't a problem is you are operating in a small range of displacement, far away from solid. On the contrary, valve springs may require many excursions to solid, therefore requiring as low of a stress at solid as possible, and depending on the constraints imposed, titanium can be significantly lower in stress and mass.


3. Temperature can be an important consideration, especially for valve springs. Cr-Si steel begins to experience setting (creep) and relaxation at temperatures above 150 C. Titanium Beta-C alloy will retain a higher percentage of its strength at this temperature.

[Toby Padfield]

Please disregard the two -- they appear due to my use of a colon after a right paranthesis...apparently I should have previewed before submitting.

[desmo]

Good info there Toby ,I've had the same thing happen. There is an option on the Reply page to disable smilies, but I've never used it.

[Top Fuel F1]

Re: Some of the many definitions for some associated words: Set = To go down; sink; wane, Setting = Fixed into a position by an outside force, Settling = the act of a thing that settles; sinking down.

I simpler times I can recall leaning against the corner of an 18 wheeler and watching Austin Coil (John Force's crew chief) overhaul the F/C heads between rounds. The time between rounds were a lot longer then, especially on Qual. days. The springs were some kind of steel alloy and the crew did a lot of work outside, in that they didn't have multiple trucks like now. He would completely disasemble the head and check each valve spring assembly in a lever operated dial gauge rig, looking for ones that had gone out of Spec. during the last run.

In a late 1998 publication Dale Armstrong ( Kennny Burnstein's crew chief at the time ) related: That they put in 1 Lb of air to 2 lbs of 100% Nitro. They are actually burning at about 1-to-1. So the unburnt 1 pound of fuel is used as a coolant. The HP comes from high cylinder pressure. However the tremendous internal heat and cyl. pressure produced would melt the engine without the cooling by the fuel. Nitro burns slowly having a far greater expasion ratio than other fuels and the peak pressure essentially continues through out the power stroke. There is still a lot of residual cylinder pressure and unburned fuel remaining when we open the exhaust valves.

There is definitely a reason for the TF & FC guys going to Ti valve springs. The crew chief doesn't have to be a metallurgist to understand what he sees on the dial guage or for that matter in out right valve spring failures. Also reseaching what some might say about what happens to Ti valve springs in F1, CART and IRL may not directly translate into what happens to them in a TF type engine, having a much more hostle environment.