The chassis con
Discussion
Avocet said:
But the first photo in the "spaceframe" section of the same article is of a Tuscan chassis. So I guess even the author of this article is confused! Am I right in thinking that Mark B is saying that the crucial element of a spaceframe is that it has to have tubular side members?
There is a technical description of a “space frame”. In layman’s terms it’s a structure in which all the structural members are in either compression or tension but NOT in twist or sheer. Somebody correctly said earlier that such a structure would work if all its joints were rose joints. Steel is very strong in both tension (look a what you can pull with a steel cable) and will support enormous loads under compression (typical pit prop). However in sheer steel is quite ‘floppy’. If you pick up a full length (7.6m) of 50x100 box section it will bend by al least 60mm under its own weight. Cut it in half and join it with cross-members and that’s what’s called a ladder chassis. They can and do bend every which way with the engine, road and cornering forces to which they’re exerted.
Don’t worry though
. That’s the way it is, but space frame technology can improve the situation
Look at a simple jib. Hang something heavy on the end and the jib bends downward. Place a diagonal ‘gusset’ from near the end of the jib back to its support post and the jib won’t bend under load. You have ‘triangulated’ the structure. The original jib is no longer forced down by the load because the gusset carries the load in compression. The load on the jib is carried in tension so could be made of much lighter material.
The essence of this principle, if correctly applied to a chassis, results in a chassis that has no beams (which bend) but a number of ‘gussets’ and ‘jibs’ which form triangles of enormous strength. When a structure is so formed, it may be rightly called a ‘space frame’ irrespective as to where or what its application
. Look for example at a crane on a building site. Its main upright and jib are not generally solid, are they? To save weight they’re built up using triangles (sometimes with one side, the one exclusively under tension, being made of wire rope). The crane is still a crane, the beams it’s made of are still beams, but it’s a crane of space frame construction and its jibs are of space frame construction.
So, how would you differentiate between an ordinary solid beam and a space frame beam. The same answer applies to a solid, as opposed to a space frame backbone (which when all is said and done is only another name for a beam).
There are other issues as to when is a tubular chassis a space frame chassis. I’m not going down that route now, save to say that it’s not very often.
Den.
Avocet said:
Take a look at <a href="http://autozine.kyul.net/technical_school/chassis/tech_chassis.htm">http://autozine.kyul.net/technical_school/chassis/tech_chassis.htm</a>
It specificaly states under "Backbone Chassis"
"TVR's chassis is adapted from this design - instead of a rigid backbone, it uses a lattice backbone made of tubular space frames. That's lighter and stronger (mainly because the transmission tunnel is wider and higher). "
So Den's descriptions sounds 100% correct from this.
But the first photo in the "spaceframe" section of the same article is of a Tuscan chassis. So I guess even the author of this article is confused! Am I right in thinking that Mark B is saying that the crucial element of a spaceframe is that it has to have tubular side members?
The point I was making was purely that a description space frame backbone chassis is misleading. Hence, my point tubular backbone chassis, and indeed the author of this article has described it as lattice backbone which in my mind is similar to tubular backbone, but there you go, I suppose each to their own...
My point about spaceframes is that they are a framework around the occupied space and yes it does include side bracing, roll over hoops and screen supports.
The picture of the Tuscan is a misleading one, it uses a roll hoop, screen support and looks to include additional Truces around the engine bay which is why the author quotes it as a spaceframe. The fact it also incorporates the carry over backbone is irrelevent as it purely moves it on from 'tubular backbone' to space frame, but I think that description is loose for this design....
Ferg said:
Mark,
Our 'other car' is a Multipla.....(pause for ridicule from people who've never owned or driven one).......and it's described as being a spaceframe! How's that work??
Ferg,
Don't worry, I have one too. It's excellent and much cooler than your average Sharan/Galaxy etc IMHO.
I think the conclusion we must then come to is that there is no particular format for a "spaceframe" chassis. It could apply equally to a "backbone" or a "tub" design as long as the structural principles (no bending loads in any member) are upheld.
I think "tubular" would be the best way to describe the TVR chassis and I can see that it has a "backbone-like" structure -just as the Lotus does but made from tubes rather than sheet. I don't think I've ever seen a true "spaceframe" chassis (backbone or otherwise)because they all end up being compromised as a result of having to fit engines and people in! Certainly, the TVR wouldn't qualify as a "spaceframe" under Kitcarman's definition and it's a definition that I've seen used before. To be honest, I can't think of any tubular chassis that don't rely to some extent on the bending stiffness of some of their tubes.
Any thoughts on that anyone?
I think "tubular" would be the best way to describe the TVR chassis and I can see that it has a "backbone-like" structure -just as the Lotus does but made from tubes rather than sheet. I don't think I've ever seen a true "spaceframe" chassis (backbone or otherwise)because they all end up being compromised as a result of having to fit engines and people in! Certainly, the TVR wouldn't qualify as a "spaceframe" under Kitcarman's definition and it's a definition that I've seen used before. To be honest, I can't think of any tubular chassis that don't rely to some extent on the bending stiffness of some of their tubes.
Any thoughts on that anyone?
I've just seen the MK GT1 chassis pics.
Spot the triangulation!
Front suspension region - there isn't much!
Rear suspension region - not a lot!
Top of the engine bay - zip!
I seriously think that MK would be a lot better off copying the Lotus 23 chassis. Alternatively why not offer to help "Project LMP" (check out his website - some interesting stuff) to offset his build costs by offering to buy a licence to copy his chassis? Why not use a simpler, cheaper, stiffer amd possibly not much heavier ladder frame?
I hope MK add a few diagonals. I've actually offered ( on another forum) to take a quick look at their design using a not too dissimilar FE model I have from my earlier ground work. I reckon the basic design could be made to work a lot better.
Spot the triangulation!
Front suspension region - there isn't much!
Rear suspension region - not a lot!
Top of the engine bay - zip!
I seriously think that MK would be a lot better off copying the Lotus 23 chassis. Alternatively why not offer to help "Project LMP" (check out his website - some interesting stuff) to offset his build costs by offering to buy a licence to copy his chassis? Why not use a simpler, cheaper, stiffer amd possibly not much heavier ladder frame?
I hope MK add a few diagonals. I've actually offered ( on another forum) to take a quick look at their design using a not too dissimilar FE model I have from my earlier ground work. I reckon the basic design could be made to work a lot better.
Surely a "spaceframe" is just a 3 dimensional frame structure? In the same way a ladder chassis is a two dimensional frame structure. The fact that practical spaceframe would have all members under tension or compression is just an artifact of good design, i.e. I would still consider a completely non-triangulated, three dimensional frame as a "spaceframe", (although probably not a very good one!)
Hence in the case of the Tiv, the chassis uses a backbone constructed in a spaceframe style.
Hence in the case of the Tiv, the chassis uses a backbone constructed in a spaceframe style.
That's certainly the conclusion I'm coming to. "Spaceframe" is a type of structure, not a particular layout. The boundaries a blurred, though. Many ladder chassis are pretty 3-dimensional. My old Scimitar had something which most people would describe as a ladder chassis (including me!) but it was at a higher level over the rear axle. (still as stiff as an over-cooked piece of sphagetti, mind)! It would then make sense (IMO) to call a TVR a spaceframe with a backbone (or a backbone spaceframe if preferred) and most 7s would be a "tub" spaceframe. The fact that neither are true "spaceframes" is probably largely irrelevant if most people think of a spaceframe as something made of tubes. We could drop the word "spceframe" altogether and just say "tubular" in fact - that might satisfy the most fervent purist! If the definition that I would prefer is adopted, probably only tower cranes would qualify as spaceframes anyway!
Mr2Mike said:
i.e. I would still consider a completely non-triangulated, three dimensional frame as a "spaceframe"
I'd agree with you on this. A simple rectangular box-shaped frame (chassis?) could be created with 12 tubes or bars around the perimeter. One could call that a spaceframe. Admittedly, it would not be a good one, although adding triangulation across each rectangle would help, but that isn't the point.
It seems there is a bit too much discussion on terminology here, since interpretations can always vary a bit - whatever book or article you choose to use for reference.
The concept of a 'ladder frame' being generally 2-D and a 'spaceframe' being 3-D makes sense to me though. Both may use tubular contruction, or may have bulkheads and panels affixed to improve strength and rigidity.
And is a (hollow) rectangular bar also a tube? Oops, getting into more terminology issues here I guess!
This is a really tough thread, isn’t it?
It started with the proposition that it’s a con to suggest that spaceframe chassis are always better, in terms of torsional rigidity, than ladder frame chassis.
My observation is that the proposition is true in practice because many of those chassis which are described as spaceframe are not, but are rather simply made out of tubes arranged three-dimensionally
.
I disagree that a three-dimensional tubular chassis can automatically be described as a space frame. Look for example at a conventional chair. It’s four legs support vertical loads in compression, but the lateral loads on those same legs place them in sheer. The back supports are always in sheer. So, as a three-dimensional tubular structure it’s most definitely not a spaceframe, but it performs admirably.
So it is with vehicle chassis IMHO. However, in the case of a chassis, the loads to which it’s subjected are infinitely more complex and are far greater.
The really sad part
is that I sincerely believe that many tubular chassis could be made into spaceframes without significant increased production costs. Or in certain cases could be made into far better structures by slight alterations in the positions of certain tubes.
To put this differently, there are so-called spaceframe chassis that are simply not spaceframes. There are others which have elements of spaceframe design, but unfortunately the designer hasn’t truly grasped the principles of triangulation to use the strength of steel axially. There are yet others where the designer has consciously used elements of spaceframe design on parts of the chassis and reverted to simple beams in other places.
My belief is that a true spaceframe will always offer a better stiffness to weight ratio than a ladder chassis. However, I agree that most chassis which are described as spaceframe are not very well designed, and therefore they do not offer the stiffness advantage that they could. Neither do they necessarily offer their potential weight saving either.
So, my advice is to accept that all three-dimensional tubular chassis are spaceframes BUT to accept that there’s a world of a difference between a good design and a bad design of spaceframe. In other words, don’t simply assume that a tubular chassis is automatically better than a ladder chassis.
Den 
It started with the proposition that it’s a con to suggest that spaceframe chassis are always better, in terms of torsional rigidity, than ladder frame chassis.
My observation is that the proposition is true in practice because many of those chassis which are described as spaceframe are not, but are rather simply made out of tubes arranged three-dimensionally
. I disagree that a three-dimensional tubular chassis can automatically be described as a space frame. Look for example at a conventional chair. It’s four legs support vertical loads in compression, but the lateral loads on those same legs place them in sheer. The back supports are always in sheer. So, as a three-dimensional tubular structure it’s most definitely not a spaceframe, but it performs admirably.
So it is with vehicle chassis IMHO. However, in the case of a chassis, the loads to which it’s subjected are infinitely more complex and are far greater
. The really sad part
is that I sincerely believe that many tubular chassis could be made into spaceframes without significant increased production costs. Or in certain cases could be made into far better structures by slight alterations in the positions of certain tubes. To put this differently, there are so-called spaceframe chassis that are simply not spaceframes. There are others which have elements of spaceframe design, but unfortunately the designer hasn’t truly grasped the principles of triangulation to use the strength of steel axially. There are yet others where the designer has consciously used elements of spaceframe design on parts of the chassis and reverted to simple beams in other places. My belief is that a true spaceframe will always offer a better stiffness to weight ratio than a ladder chassis. However, I agree that most chassis which are described as spaceframe are not very well designed, and therefore they do not offer the stiffness advantage that they could. Neither do they necessarily offer their potential weight saving either.
So, my advice is to accept that all three-dimensional tubular chassis are spaceframes BUT to accept that there’s a world of a difference between a good design and a bad design of spaceframe. In other words, don’t simply assume that a tubular chassis is automatically better than a ladder chassis.
Den
There is another point to be made as well, some people seem to be referring to ladder frames when they mean back bone chassis... This is again an error, a backbone chassis is not the same as a ladder frame. To understand the difference, as usual typical applications of both:
Back bone, as mentioned TVR or GD Cobra....
Ladder frame - Land Rover, Toyota Hi Lux, Mitsubishi L200.
These are fundamentally flawed in torsional rigidity and hence they are used in applications where torsional rigidity is not so important, and in fact better that it is less stiff. ie 4x4 style vehicle where chassis twist aids the cross axle wheel travel..
Back bone, as mentioned TVR or GD Cobra....
Ladder frame - Land Rover, Toyota Hi Lux, Mitsubishi L200.
These are fundamentally flawed in torsional rigidity and hence they are used in applications where torsional rigidity is not so important, and in fact better that it is less stiff. ie 4x4 style vehicle where chassis twist aids the cross axle wheel travel..
That's possibly a slight over-simplification I feel! There have been a number of cars where a "ladder" chassis (or even a "perimeter frame") has been used but stiffness is still important. Some manufacturers prefer to use the shell for the torsional stiffness and keep a simple chassis merely to hang suspension & drivetrain components on. Effectively the chassis is little more than a couple of longitudinal members holding front and rear subframes together.
jchase,
I did the analysis using the longitudinal axis of the car. I fixed the rear suspension spring mounts and applied loads to the front spring mounts to give a twisting moment. From the applied moment and the displacement results I calculated the chassis stiffness. My models only include the welded steel tubes and panels. It is difficult to assess the impact of riveting alloy panels to a steel frame as while this can make a big difference when new the rivet work loose over time. The panels may appear to be secure but their contribution to stiffness may be greatly reduced. Bodyshells can also enhance stiffness but modeling a composite structure aswell as the chassis would make the models very complex.
What nakes you ask by the way? It sounds as if may be doing some FE of your own!
I did the analysis using the longitudinal axis of the car. I fixed the rear suspension spring mounts and applied loads to the front spring mounts to give a twisting moment. From the applied moment and the displacement results I calculated the chassis stiffness. My models only include the welded steel tubes and panels. It is difficult to assess the impact of riveting alloy panels to a steel frame as while this can make a big difference when new the rivet work loose over time. The panels may appear to be secure but their contribution to stiffness may be greatly reduced. Bodyshells can also enhance stiffness but modeling a composite structure aswell as the chassis would make the models very complex.
What nakes you ask by the way? It sounds as if may be doing some FE of your own!
cymtriks said:I thought that on most kits (certainly Westfield and Ultima) the panels are bonded in addition to the rivets.
It is difficult to assess the impact of riveting alloy panels to a steel frame as while this can make a big difference when new the rivet work loose over time. The panels may appear to be secure but their contribution to stiffness may be greatly reduced.
Cymtrics and BobM,
You’re both right and you’re both wrong, I reckon.
The panels that would predominantly stop twist are those arranged across the width of the car. Have you seen the advert for a windscreen company that claims “your windscreen can account for 30% of your cars strength”?
Thus proving that a bonded in panel can and does contribute to stiffness (rivets not recommended for windscreens). I reckon that (if true) would prove Bob right.
However, in the kind of cars were talking about the front and rear bulkheads would perform similarly. Excepting, there are few instances where there’s a single bulkhead panel
. Usually such panels are cut in half (or severely interrupted) by the transmission tunnel, in the case of a rear bulkhead, or the gearbox bulge in the case of a front bulkhead. How these two halves are ‘joined’ or ‘linked’ or ‘braced’ by the underlying chassis structure is THE underlying means through which the structure shall prove torsionally stiff. The presence of the gearbox (at the front) or propeller shaft (at the rear) usually means there’s little or no such cross-bracing at the lower level. Result is that instead of one stiff bulkhead, there are two (almost) unconnected stiff bulkheads which amounts to one un-stiff bulkhead.
My point is that the overall structure won’t be significantly strengthened by bonding metal sheets onto the two bulkhead halves. Thus, I’d say that Cymtrics is right, but for a different reason to the one he’s given.
Den
PS Heavy thread Man
You’re both right and you’re both wrong, I reckon. The panels that would predominantly stop twist are those arranged across the width of the car. Have you seen the advert for a windscreen company that claims “your windscreen can account for 30% of your cars strength”?
Thus proving that a bonded in panel can and does contribute to stiffness (rivets not recommended for windscreens). I reckon that (if true) would prove Bob right. However, in the kind of cars were talking about the front and rear bulkheads would perform similarly. Excepting, there are few instances
where there’s a single bulkhead panel. Usually such panels are cut in half (or severely interrupted) by the transmission tunnel, in the case of a rear bulkhead, or the gearbox bulge in the case of a front bulkhead. How these two halves are ‘joined’ or ‘linked’ or ‘braced’ by the underlying chassis structure is THE underlying means through which the structure shall prove torsionally stiff. The presence of the gearbox (at the front) or propeller shaft (at the rear) usually means there’s little or no such cross-bracing at the lower level. Result is that instead of one stiff bulkhead, there are two (almost) unconnected stiff bulkheads which amounts to one un-stiff bulkhead. My point is that the overall structure won’t be significantly strengthened by bonding metal sheets onto the two bulkhead halves. Thus, I’d say that Cymtrics is right, but for a different reason to the one he’s given.
Den
PS Heavy thread Man
kitcarman said:Last time I looked my Ult didn't have a transmission tunnel or gearbox bulge
Excepting, there are few instances
Wasn't I supposed to hang the gearbox off the back? 
Seriously though, surely any panels will increase the torsional strength if they cover a whole 'triangle' or square even? I do of course accept that if one spans several triangles or squares it'll be of more benefit.
BobM said:
Last time I looked my Ult didn't have a transmission tunnel or gearbox bulge
That’s one of the reasons why an Ultima is such a good product. Mid engine designs do enable a ‘purer’ structure to be created. Skinning a good structure in aluminium sheet makes it better. BobM said:Sure, the cladding of individual parts of any structure will strengthen those parts. The point I was attempting to make is that if the structure isn’t right in the first place, that is to say if it’s parts aren’t joined correctly, then the ally sheet skin will not generally significantly improve the structure as a whole.
Seriously though, surely any panels will increase the torsional strength if they cover a whole 'triangle' or square even? I do of course accept that if one spans several triangles or squares it'll be of more benefit.
Look at it this way. Imagine the two halves of the bulkhead like two railway goods box cars joined together by a coupling. They will move relative to each other. Making each stronger (by panelling their sides for example) will not stop them moving relative to each other
. A good space-frame chassis design will seek to avoid having two separate structures by replacing the ‘coupling’ (the weak link) such that the two box cars in this analogy become like one box car but twice the length. Clearly, any increase in the stiffness of this longer box (by panelling for example) will obviously lead to a directly proportional increase in the strength of the overall structure. In the case of a chassis which is made of two halves separated (or severely weakened
) by a transmission tunnel then sheeting the bulkheads either side of the weakness won’t help very much, if at all
. There are numerous solutions. Ultima gets rid of the tunnel. GD work on the tunnel itself to make it THE structural member thus negating the need for structurally important bulkheads. Westfield/Caterham run structural members under the tunnel, such as to have the effect of a structural continuous bulkhead, but not its appearance. The vast majority of kit cars revert to heavyweight beams
traversing what might otherwise have been a much lighter space-frame design of chassis. There are yet other (so called) designs which utilise heavy beams everywhere in what to the uninformed might otherwise look like a space frame. Does any of this matter? Not unless you’re racing. At the end of the day ladder frames work and it’s hardly possible to make worse structure in a three dimensional design (but it is done
). If it works well and it’s light, it’s probably a space frame. Otherwise it’s probably just a 3D tubular chassis. Either way, for road use
it’ll probably do you proud. Den
. OK, just to take things a step further...
I understand Kitcarman's arument about the two rigid bodies joined by a non-rigid structure. The result is still a non-rigid structure.
BUT, it's all very well analysing a car by restraining its rear spring mounts and applying a torque to the front ones, but, strictly speaking, that's not how they're twisted on the road. To my mind, you need to put the wishbones, hubs, wheelbearings, & balljoints on and then fit "lockup" bars in place of the springs. I think the chassis should then be restrained by securing the rear hubs and loaded by applying the torque through the front hubs each in the same vertical plane as the centre of the wheel. This, after all, is how they're loaded on the road. The earlier argument can then be taken a step further to say "not only is it pointless having two rigid bodies connected by a floppy structure, but its also pointless having a rigid structure connected to the road by bendy wishbones and squidgy bushes.
I understand Kitcarman's arument about the two rigid bodies joined by a non-rigid structure. The result is still a non-rigid structure.
BUT, it's all very well analysing a car by restraining its rear spring mounts and applying a torque to the front ones, but, strictly speaking, that's not how they're twisted on the road. To my mind, you need to put the wishbones, hubs, wheelbearings, & balljoints on and then fit "lockup" bars in place of the springs. I think the chassis should then be restrained by securing the rear hubs and loaded by applying the torque through the front hubs each in the same vertical plane as the centre of the wheel. This, after all, is how they're loaded on the road. The earlier argument can then be taken a step further to say "not only is it pointless having two rigid bodies connected by a floppy structure, but its also pointless having a rigid structure connected to the road by bendy wishbones and squidgy bushes.
Gassing Station | Kit Cars | Top of Page | What's New | My Stuff