2x4 frames vs spaceframes

[Jean-Sebastien Stoezel]

Most (if not all) the FV I've seen have a 2x4 (or 1x4) bottom frame. You don't seem to see similar structures in other classes, like FF, FC etc... Anybody knows why this structure is so specific to FV?
The 2x4 structure seems quite heavy, and not that rigid torsion-wise. You would think spaceframes would provide much more rigidity.

I read on a FV historic website that apparently the 2x4 bottom was introduced with the Lynx B, because it was easier to manufacture and fix. Why have builders stick to that while other classes have been using spaceframes? ? Is it also that the 2x4 frame is much easier to couple the VW engine/transmission to the chassis?

FV series in Australia and England all use spaceframes, the 2x4 structure seems to be typical to North America.

[problemchild]

The ladder-style chassis is commonly used when torsional rigidity is not critical (as with zero-roll FV suspension). Its main attraction is the low cost of production. Almost anybody with a mig welder, saw, and grinder can build a chassis in a very short time period.

[Jean-Sebastien Stoezel]

The ladder-style chassis is commonly used when torsional rigidity is not critical (as with zero-roll FV suspension). Its main attraction is the low cost of production. Almost anybody with a mig welder, saw, and grinder can build a chassis in a very short time period.

You would think torsional rigidity is always important, even more for a zero roll suspension, since only the front is responsible for keeping the car from rolling over (with the anti roll bar). And all that force transfer somehow has to go through the whole length of the frame. Hence the need for a rigid frame.

[Hardingfv32]

1) Torsional rigidity is important when there is a relationship between the front and rear suspensions in roll. There is no such relationship with zero roll. With the front doing all the roll control, what is the difference if the rear half of the car twists a few minutes of a degree? Rear camber will be affected very slightly.

2) A FV only has a few mounting hard points to accommodate: front beam, front and rear of gearbox on most FV's. The ladder frame does this just fine and also throws in low CG and side protection. A normal formula car has 8 front suspension mounting points, 2 shocks/rockers systems and possibly as many rear points to accommodate. The ladder frame also holds up better in accidents and is probably easier to repair.

Brian

[Bruce Livermore]

The of the main reasons chassis stiffness is important on almost all cars is that it allows tuning to be done by adjusting the relative roll resistance between front and rear primarily by using anti-sway bars. Since a "zero roll" Vee puts all the roll resistance at the front, by definition we don't (typically) use anti-sway bars to tune our car's handling, at least not the steady state handling. There are other reasons to want chassis stiffness and it probably never hurts, but many successful Vees have been ladder frames. As Greg says they are easy and cheap and that suits FV very well.

Edit: Yeah, Like Brian said mostly ;-)
Got distracted...

[Jean-Sebastien Stoezel]

The of the main reasons chassis stiffness is important on almost all cars is that it allows tuning to be done by adjusting the relative roll resistance between front and rear primarily by using anti-sway bars. Since a "zero roll" Vee puts all the roll resistance at the front, by definition we don't (typically) use anti-sway bars to tune our car's handling, at least not the steady state handling. There are other reasons to want chassis stiffness and it probably never hurts, but many successful Vees have been ladder frames. As Greg says they are easy and cheap and that suits FV very well.

Edit: Yeah, Like Brian said mostly ;-)
Got distracted...

Thanks for the clarifications, that was kind of counter intuitive, or at least I found it was counter intuitive. Are most modern cars using 1x4 or 2x4 now? My old Lynx has 2x4 and they seem like overkill for it.

[Hardingfv32]

2 X 4 .120 wall is best for accident strength IF you can tolerate the weight. All the smaller dimension tubing sizes or thinner walls are just a compromise between accident strength vs weight. I'm not concern about the driver in an accident, but how well the car holds it's alignment. The designer has to make a decision what criteria he is going to use for the chassis design. I think the Vortech is 2x2. I would say a 1" X .060 tube space frame as the lightest and most costly choice.

Brian

[Ananth K]

I can't comment on the benefits of one frame construction over another, but I can say that torsional rigidity is important even if the car is equipped with zero-roll suspension.

The reason for this is that even though the so-called "zero-roll" suspension has zero roll stiffness in the back, it will still transfer load to the outside tire through the suspension links (because of the high roll center). If there were no load transfer at the rear axle your cars would probably plow like a garden tractor and be generally undrivable.

So the total loads are still the same, just that they are being fed into a different spot on the frame. Ergo, axle-to-axle torsional rigidity is still required in the same magnitude, but the optimal frame member placement might be different to achieve that rigidity.


Also as a general comment on frame construction, a 2x4 ladder-type frame need not necessarily be less torsionally rigid than a spaceframe, but it will be heavier and cheaper to make for sure.

[Hardingfv32]

I think this could be wrong.

1) There is no variable in the weight transfer formula for deg or angle of roll. I am assuming load transfer is weight transfer.

2) It would seem to me that there are very few components or even the chassis in the rear of a std FV that can flex very much during normal cornering in regard to rear weight transfer.

3) The chassis has a fixed rate of twist or stiffness. This should not change. Even so, how would weight transfer be effected? Maybe delayed by some small fraction of a second? Wouldn't this delay always be the same and thus part of the car's characteristics.

Brian

[Purple Frog]

Interestingly, Jay Novak is using multiple pieces of rectangular tubing in his newest F600 chassis. I didn't have a camera at Road Atlanta to document, but the chassis looks like a tank in terms of driver protection, and... he has to add ballast.

[Ananth K]

I think this could be wrong.

1) There is no variable in the weight transfer formula for deg or angle of roll. I am assuming load transfer is weight transfer.

2) It would seem to me that there are very few components or even the chassis in the rear of a std FV that can flex very much during normal cornering in regard to rear weight transfer.

3) The chassis has a fixed rate of twist or stiffness. This should not change. Even so, how would weight transfer be effected? Maybe delayed by some small fraction of a second? Wouldn't this delay always be the same and thus part of the car's characteristics.

Brian

hi brian,

1) there are two components of load (or weight) transfer, one elastic (from springs, bars and tires) and one kinematic. The elastic component is dependent on roll stiffness which is expressed in lb-ft per deg of roll. The kinematic component is dependent on roll center height.

2) Cant comment as I'm not familiar with Vees.

3) The whole point of roll stiffness is to control how much load is transferred in the front vs. the rear, which along the kinematic load transfer ultimately determines the balance of the car (understeer). Torsional rigidity is ultimately a measure of axle-to-axle stiffness of the chassis (the basis of the chassis being called "the third spring" in series with the front and rear roll stiffnesses). When you put these two facts together it is easy to see that a chassis with insufficient torsional rigidity will be impossible or at best difficult to tune accurately via controlling the load transfer distribution front-to-rear.

Hope this was a better explanation.

[Hardingfv32]

What I THINK...

A given car in a given turn has a fixed amount of total weight transfer. A portion of that total is weight transfer is associated with the roll centers or roll axis. This portion can be assigned between the front and rear suspension. Is this called lateral load load distribution? Roll does not create weight transfer, it is caused by weight transfer.

To modify the lateral load transfer between the ends of the car you need a roll couple. There is no roll couple with a true FV zero roll suspension. So I am assuming that there is no way to adjust the amount of weight transfer front to rear. There is no rear sway bar with a true zero roll and the rear springs only maintain the ride height.

That being the case, how does chassis stiffness matter with a zero roll FV? I build my own cars and have absolutely no concern for chassis stiffness. I'm very interested to know if I'm wrong.

Brian

[Ananth K]

Chassis stiffness matters because the chassis is still seeing loads from the jacking forces created by the high roll center in the back, even though there is no sway bar or spring that creates a torque on the frame. These loads are obviously asymmetric left to right when the car is cornering, and all I'm saying is that this has a similar effect on the frame as having roll stiffness at that end of the car.

[Hardingfv32]

I do not understand your last thought. I do not see how the jacking force extends beyond the the front or rear suspension system that it is affecting. Isn't jacking force directly associated with a particular suspension system, not chassis/frame configuration?

I do realize that weight transfer is effect by the changing height of the chassis during jacking.

Brian

[smsazzy]

Is the jacking force creating roll transfer?

Perhaps. In a right hand turn, the outside (left rear) tire is pushing up on the rear suspension push rod. Some of that force tries to lift the rear of the car, some of that force is transferred over and through the spring to push the inside tire back down.

If this is what is happening, then some of that jacking force, could be transferred through the chassis as torsional "twist" of the frame.

Am I on the right track, or way off base?

[Hardingfv32]

The jacking force is purely vertical, working against the weight of the car at the rear suspension location.

Now, you could propose that the chassis could flex (sag) as it is lifted, but what would be the harm? Remember the jacking action is well controlled by the rear shock's rebound valving. It is very unlikely you are going to be able to detect any chassis sagging in this case.

Brian

[Ananth K]

I do not see how the jacking force extends beyond the the front or rear suspension system that it is affecting. Isn't jacking force directly associated with a particular suspension system, not chassis/frame configuration?

Consider this: Lets say you have roll center height = 0 (on the ground) at the rear of the car. Then all the load transfer at that end of the car is happening from the springs and ARB. So you have X amount of torque going into the chassis that is resisted by the torsional stiffness of the frame.

Now lets say you take all the roll stiffness out of the rear, and raise the roll center height to the point where the TOTAL load transfer at the rear is taken care of SOLELY because of the jacking force caused by the roll center. Since the steady-state balance of the car has not changed because the load transfer distribution is the same, then there still must be X amount of torque going into the chassis, and that still has to be resisted by the torsional stiffness of the frame.

Is this explanation helpful?

[Hardingfv32]

The explanation is not helpful, but it could be some preconceptions I have on the subject. Let's keep hammering away at it. Trying to learn, not win the discuss.

First, I will contend that jacking has nothing to do with load (weight) transfer .. IF.. we disregard the fact that the CG has been raised. That is not what we are interested with in this discussion.

So, if the car jacks then the roll center (and roll axis) move up the same amount as the CG. I do not think anything has change. All the sprung weight that the rear suspension is allocated is transfer thru the high roll center. The chassis might be pivoting above roll axis at the rear, but with no resistance being provided, no torque can be generated that might cause the chassis to twist.

As a physical note, two people can pickup the front end of a zero roll FV and rotate it 15-20 deg either direction with all most no effect on the rear tires.

Brian

[Ananth K]

The jacking force on the chassis is directly proportional to three things: the lateral force generated at the contact patch, the roll center height and the track width.

So for a given vehicle, as you increase lateral force at the contact patch, two things happen, assuming there is both roll stiffness and a roll center above ground.

1) A jacking force is generated which is proportional to the lateral force. This jacking force is transmitted to the chassis via the suspension links. You are right in that the jacking force is purely vertical, but this is only true AT THE CONTACT PATCH. The orientation of the forces and the moments it puts on the chassis depends on how the suspension links are arranged.

2) An inertial force proportional to the sprung mass of the vehicle acts through the CG, which causes the vehicle to develop a roll angle, and this is transmitted to the chassis (or reacted by the chassis, however you want to think about it) via the springs and ARB.

What I am saying is that, even if the force component described in #2 above equals zero (i.e. in a zero-roll FV), the force component in #1 still exists and still needs to be reacted by the chassis, therefore, I think chassis stiffness is still important when one end of the car doesnt have any roll stiffness.

Cant make it any more simpler than that

[Hardingfv32]

So, in reference to statement #1, the lateral force acting on the center of the chassis must make its way to the rear and front suspension. If the chassis flexes laterally, then how is this going to manifest itself? How is the handling going to be affected? Will less lateral force make it to the rear? Less jacking?

Maybe some actual cause and effect might help me understand.

Brian

[Ananth K]

OK, here are some cause and effect statements; The analysis is greatly simplified for clarity:

lateral force starts to develop at the contact patches (cause) and REACTED at the CG as an inertial force (effect).

The lateral forces at the contact patches cause different jacking forces front to rear (because roll center heights are different, obviously)

The jacking forces are transmitted to the chassis via the suspension links (cause), and the effect is obviously that the chassis must absorb these forces via its torsional stiffness.


Now at this point if you were to assume that the chassis had absolutely no springs or ARBs at all, much like a kart, then you'd see that chassis torsional stiffness is absolutely important in reacting to the jacking forces. In a kart (a proper race kart, not a concession go-kart), the chassis is intentionally made compliant torsionally, probably to absorb some road inputs, and partly to ensure that the inside rear wheel lifts to allow the solid-axle a pseudo-differential effect in tight turns found on a typical kart track.

On the other hand, if the vehicle in question has springs and ARB, then the moment created by the CG being above the roll axis causes it to roll on its springs, and since the roll stiffnesses front to rear are almost always different, this difference in roll stiffness is what needs to be reacted by the torsional stiffness of the chassis.

In a right hand turn, the outside (left rear) tire is pushing up on the rear suspension push rod. Some of that force tries to lift the rear of the car, some of that force is transferred over and through the spring to push the inside tire back down.

If this is what is happening, then some of that jacking force, could be transferred through the chassis as torsional "twist" of the frame.

Sorry, I completely missed this. You hit the nail right on the head. This is what I've been trying to say all along in poor engineerese.

[Hardingfv32]

OK, I think I have this figured out. My thoughts on the subject of twisting torque being applied to the rear of the chassis:

I am assuming we are talking about the jacking force only at the rear. The jacking force is applied to the chassis where the axle pivots in the transmission and to a much lesser amount where the control arms attaches to the chassis. The axles are 4.75" apart in the transmission and for this discussion I assume the control arms are also 4.75" apart (leading arm mounted at very rear of the chassis).

The following numbers come from a spread sheet John Petillo is developing regarding FVs:

The two swing axle rotation points are separate by about 4.75". Let's consider the two swing axle rotation points being held apart by 4.75" "bar" from side to side. The two torques pushing are caused by the jacking force. Effectively the outer wheel pushes up and the inner wheel pulls down. For our -2 degrees camber and 1.4G cornering force, jacking causes 278 lbs up on the outer end of the "bar" and 62 lbs down on the inner. The difference in these is the jacking force of 216 lbs up. But these forces both contribute to a torque on that "bar" of (278-(-62))*(4.75/2) = 808 lb-in = 67 lb-ft of twisting torque on the frame. Not sure at this point what small amount of torque might be generated by the control arm mounting points. I suspect it just represents a portion of the number stated above, not anything additional.

It should be noted that the torque due to jacking forces resists the roll moment of the chassis. It acts in the opposite direction, so it reduces the torque that the chassis has to deal with.

The action of the rear spring and rocker system (zero roll mechanism) is a little more complex and not fully understood at this point. It should be noted that the rear spring is in rebound and reducing the force it provides as the rear suspension jacks. There could be an difference side to side at the mounting points of the rockers in how this force reduction is applied to the chassis. We are evaluating this now.

So... If I ASSUME that the torque to the rear of the chassis is 67 lb-ft, this would be the number that my chassis must absorbed without twisting, correct? This I can test for on the shop floor.

Brian

[Ananth K]

The two swing axle rotation points are separate by about 4.75". Let's consider the two swing axle rotation points being held apart by 4.75" "bar" from side to side. The two torques pushing are caused by the jacking force. Effectively the outer wheel pushes up and the inner wheel pulls down. For our -2 degrees camber and 1.4G cornering force, jacking causes 278 lbs up on the outer end of the "bar" and 62 lbs down on the inner. The difference in these is the jacking force of 216 lbs up. But these forces both contribute to a torque on that "bar" of (278-(-62))*(4.75/2) = 808 lb-in = 67 lb-ft of twisting torque on the frame. Not sure at this point what small amount of torque might be generated by the control arm mounting points. I suspect it just represents a portion of the number stated above, not anything additional.

Looking at your loads of 278 and 62 lbs I presume that you've taken load transfer into consideration? In other words you're reducing the lateral force applied at the inner and increasing the lateral force applied at the outer patch? If you have then it certainly looks reasonable.

It should be noted that the torque due to jacking forces resists the roll moment of the chassis. It acts in the opposite direction, so it reduces the torque that the chassis has to deal with.

In a Vee, yes, but not in a vehicle where the roll center is below the ground plane.

So... If I ASSUME that the torque to the rear of the chassis is 67 lb-ft, this would be the number that my chassis must absorbed without twisting, correct? This I can test for on the shop floor.

No, the correct number would be the torque due to combined effects of front roll stiffness and jacking, minus the torque due to the same combined effects at the rear.

It has been suggested that a good torsional stiffness number to shoot for is 10 times the difference in roll stiffness front to rear, in cars where the contribution from the jacking is somewhat negligible.

[hysteria]

Anybody has any idea how heavy a FV frame should be for minimum weight, that is the bare frame, with or without the bottom panel?

2 X 4 .120 wall is best for accident strength IF you can tolerate the weight. All the smaller dimension tubing sizes or thinner walls are just a compromise between accident strength vs weight. I'm not concern about the driver in an accident, but how well the car holds it's alignment. The designer has to make a decision what criteria he is going to use for the chassis design. I think the Vortech is 2x2. I would say a 1" X .060 tube space frame as the lightest and most costly choice.

Brian

[Mark Filip]

That would depend on the driver size.

[samiam520]

My Crusader FV chassis weighs 145lbs with the belly pan. I weigh 200-210lbs with gear on. Car in race trim comes in around 1050lbs

[hysteria]

Whoa, while this seems like a heavy frame (most FSEA are sub 80lbs) it's interesting you are still very close to minimum weight...

What is the regulation for the belly pan on a FV? I looked in the GCR and could not find the information. If I missed it and somebody could point it to me that would be great.

Do you need to have a welded support frame underneath the driver? Do you need a steel belly pan?

I've heard of people having a steel plate underneath the driver, then having a full aluminum undertray. I've also heard some belly pan are glued...

My Crusader FV chassis weighs 145lbs with the belly pan. I weigh 200-210lbs with gear on. Car in race trim comes in around 1050lbs

[smsazzy]

I don't think FSAE frames are typically ten feet long and built to go 100 MPH wheel to wheel with another car. Most are built for autocrossing and are not as concerned with the added bulk that comes along with that.

[problemchild]

My car (http://www.apexspeed.com/forums/show...257#post278257)
has a chassis weighing 120 lbs including floor. Cars with rearward trailing arms tend to be 15-20 lbs heavier than cars with forward arms ..... although obscene numbers like 35-40 lbs are often quoted.

It is my belief that some of the top cars are under 100 lbs (with rearward arms) but I don't have the balls to drive a car with so little integrity or side protection.

I had an earlier car (Shirley Maclane), which had a chassis at 85 lbs (forward arms). Everytime it got crashed, it needed a front clip and the driver was lucky to limp away. I revised, it as I was unconfortable with the safety or crashability, and the final version frame was over 100 lbs.

[Matt King]

Most FSAEs I have seen would be death traps on any road course, if they could even pass SCCA tech (probably not), so it's a bad comparison.

[hysteria]

Alright, FSAE may have been a bad example.

Let me rephrase and use a different reference then, based on this thread:

http://www.apexspeed.com/forums/showthread.php?t=26988

Most FSAEs I have seen would be death traps on any road course, if they could even pass SCCA tech (probably not), so it's a bad comparison.

[smsazzy]

"Mike's numbers should be right for frames up until 01 when the frame stops behind the rollbar."

Because these frames use the engine/trans as part of the chassis, this is not included in the weight.

If you cut a Vee off at the rear rollbar, you'd be well under 100 pounds as well.

[hysteria]

Thanks for the detailed answer, that's a nice car you've built. Any info on how what the rules specify for the belly pan?

My car (http://www.apexspeed.com/forums/show...257#post278257)
has a chassis weighing 120 lbs including floor. Cars with rearward trailing arms tend to be 15-20 lbs heavier than cars with forward arms ..... although obscene numbers like 35-40 lbs are often quoted.

It is my belief that some of the top cars are under 100 lbs (with rearward arms) but I don't have the balls to drive a car with so little integrity or side protection.

I had an earlier car (Shirley Maclane), which had a chassis at 85 lbs (forward arms). Everytime it got crashed, it needed a front clip and the driver was lucky to limp away. I revised, it as I was unconfortable with the safety or crashability, and the final version frame was over 100 lbs.