29 replies to this topic

[mariner]

Looking through my old Motor Sport copies I came across a group track test of 1970's F3 cars. Chevron, Brabham and Lotus 59. The writer made the point about the 59 having a square tube frame to ease production.

That made me think, was the 59 the first formula car ( i.e built for an approved FAI, SCCA or RAC formula) to use square chassis tubes.

I kind of think it was but not sure

[Peter Morley]

March 693 apparently had a square tube space frame and that also appeared in 1969.

[Catalina Park]

My 1969 MJM Roadrunner Formula Vee has some square tubes in its chassis. The lower main tubes start off square then taper to rectangular, the upper tubes are round and some of the upright tubes are square.
I seem to recall some others from the same era with square tubes.

Didn't the Lister F2 have square tubes? So that would take it back another decade.

[Tim Murray]

I think that most (all?) Mallocks had square-tube spaceframes.

[macoran]

March 693 apparently had a square tube space frame and that also appeared in 1969.

Marchives lists 703 as having a spaceframe based on 693..............and the 703 spec sheet says it has a square tube spaceframe.

[RonPohl]

Didn't the Lola 200 formula ford series have some square tubes?

[Ray Bell]

The Bowin P4 had square tubes...

I've no doubt there were many cars did before that.

[Doug Nye]

Depends how you define 'square' tubes. Colin Chapman specified round-edged 'square' stock in the Lotus 12 and 16 single-seaters as early as 1956-58 and I would feel confident there are precedents for them.

DCN

[Mistron]

round-edged 'square' stock

DCN

I think this it the difference between 'box' section and 'Rectangular hollow section' - the radius on the corners. RHS of course can include 'square hollow section'

I can't recall which has the sharp corners and which is rounded though!

[Bloggsworth]

I think this it the difference between 'box' section and 'Rectangular hollow section' - the radius on the corners. RHS of course can include 'square hollow section'

I can't recall which has the sharp corners and which is rounded though!

Depends on your definition of a radius - all the square/rectangular section CDS or ERW tube (up to 14swg/2mm) I ever used had radiused corners, very small ones granted, but definitely radiused. Hollow section would not normally be used in a racing car as, IIRC, it starts at 2.5/3mm thick and is therefore a bit on the heavy side. Square tube is torsionally more rigid than round tube, and of course, far easier to jig; though whether the torsional strength/weight penalty/smaller cross-section trade off works I'm not sure, otherwise I would assume a predominance of sq. tubed chassis.

Edited by Bloggsworth, 20 May 2012 - 14:23.

[mariner]

As ever, an instant response on this forum to almost any question!

I suspect that square tubing was adopted once good quality ERW ( electric resistance welded) tube became widely available. It was so much cheaper than the cold drawn ( Reynolds etc) tube and once the designers had learnt how to do a good spaceframe with only tension and compression the expensive stuff wasn't needed.

Also weld quality and repair speed was quicker with square tube so it made sense.

[BMH Comic]

The Bowin P4 had square tubes...

I've no doubt there were many cars did before that.

Many many cars Ray, TS Special as early as 1948 and the Comic Book in 1950+ (Formula Libre) Was in common use in WA post WWII. Many fiberglass kit cars used it also. All were made from British Steel welded tube and most bus frames used the material here. I think you will find the Smith Special has some in it as well.

[D-Type]

Depends on your definition of a radius - all the square/rectangular section CDS or ERW tube (up to 14swg/2mm) I ever used had radiused corners, very small ones granted, but definitely radiused. Hollow section would not normally be used in a racing car as, IIRC, it starts at 2.5/3mm thick and is therefore a bit on the heavy side. Square tube is torsionally more rigid than round tube, and of course, far easier to jig; though whether the torsional strength/weight penalty/smaller cross-section trade off works I'm not sure, otherwise I would assume a predominance of sq. tubed chassis.

Are you sure? I have always understood that weight for weight, for a given wall thickness a circular tube is the most efficient in torsion. I'm not sure about flexural properties though. Certainly square or rectangular tubes are generally easier to fabricate as the cuts are all straight cuts, either mitres or birdsmouths.

But I have to admit I know more about concrete than steelwork.

As to the original question, I think some 500cc F3 cars had frames of square tubes.

Edited by D-Type, 20 May 2012 - 18:28.

[Bloggsworth]

Are you sure? I have always understood that weight for weight, for a given wall thickness a circular tube is the most efficient in torsion. I'm not sure about flexural properties though. Certainly square or rectangular tubes are generally easier to fabricate as the cuts are all straight cuts, either mitres or birdsmouths.

Sorry - I should have been more clear. A 1" sq tube is torsionally stiffer than a 1" diameter, hence the weight/wall thickness trade off - Is a 1" x 18swg round stiffer than a 1" sq x 20swg? Someone with knowledge and a calculator will soon admonish me for my presumption should I be wrong.

[macoran]

generally easier to fabricate as the cuts are all straight cuts

the welds will then also be straight, and straight welds have always been the weakest link i have been told
so tubular space frames ......with properly formed joints......should be stronger pound for pound

Edited by macoran, 20 May 2012 - 20:37.

[D-Type]

the welds will then also be straight, and straight welds have always been the weakest link i have been told
so tubular space frames ......with properly formed joints......should be stronger pound for pound

It could well be so as the ends and corners will act as stress inducers. It's a trade off against easier cutting.

Edited by D-Type, 20 May 2012 - 20:49.

[elansprint72]

.........

But I have to admit I know more about concrete than steelwork.

....

Basic Laws of Physics, Mr Bond; you are correct, as are D-type and Bloggs but you are all looking at the same problem from a slightly different perspective. Personally, I would use elliptical tubing, in the correct orientations and given no budgetary constraints.
A computer-thingy (without knowledge of "modern substances") would opt for square-section every time; if you asked it to include monetary considerations into the equations.

Damn, I wish I had paid more attention in those dusty rooms in that far-off World of Miracles and Magnets.

[Ray Bell]

Originally posted by BMH Comic
Many many cars Ray, TS Special as early as 1948 and the Comic Book in 1950+ (Formula Libre) Was in common use in WA post WWII. Many fiberglass kit cars used it also. All were made from British Steel welded tube and most bus frames used the material here. I think you will find the Smith Special has some in it as well.

Of course, and so did the Ranford Chrysler (1940/51), you're right about the Smith Special too...

That was larger section stuff than most of these others mentioned here, of course... 2" square or close to it I'd think.

On the subject of 'rounded corners' I recall that the Farrells had a larger-diameter rounding of the corners on the tubing in them, perhaps a quarter inch radius while the typical tube was much less so.

[mariner]

IF you twist it around its axis a round tube will be stronger torsionally than a square one of similar weight. Thats important if the chassis does twist the tube as with a twin tube chassis like a Lister or AC Cobra.

Proper spaceframes don't twist the tubes ( or shouldn't ) and each tube should only be loaded in compresssion or tension.

A square tube will have 4X the diameter, a round tube will have 3.14 ( ish) the diameter so square is about 25% heavier in the same wall thickness.

However they could use a thinner wall thickness with square , say 18 guage not 16 guage so the weght difference is not great if properly designed.

As semi monocoque chassis became more popular ( like Chevrons or the Lotus 69 ) square tube had the advantage of being much easier and better to rivet the panels onto.

[elansprint72]

IF you twist it around its axis a round tube will be stronger torsionally than a square one of similar weight. Thats important if the chassis does twist the tube as with a twin tube chassis like a Lister or AC Cobra.

Proper spaceframes don't twist the tubes ( or shouldn't ) and each tube should only be loaded in compresssion or tension.

A square tube will have 4X the diameter, a round tube will have 3.14 ( ish) the diameter so square is about 25% heavier in the same wall thickness.

However they could use a thinner wall thickness with square , say 18 guage not 16 guage so the weght difference is not great if properly designed.

As semi monocoque chassis became more popular ( like Chevrons or the Lotus 69 ) square tube had the advantage of being much easier and better to rivet the panels onto.

I believe that we are batting on the same team.

[Ray Bell]

Originally posted by marine
.....As semi monocoque chassis became more popular (like Chevrons or the Lotus 69), square tube had the advantage of being much easier and better to rivet the panels onto.

And in many spaceframe chassis of the sixties and early seventies, the bottom of the chassis was square tube while the rest was round... to facilitate rivetting on the undertray.

[Charlieman]

I more or less agree with Mariner. Costin and Phipps _Racing and Sports Car Chassis Design_ has a few notes on square section tubes. It reports that the Lola Mk1 used square section for some vertical members -- that would have eased construction for the suspension mounts and helped with the bending loads that they applied.

Even with the finest spaceframe, it is impossible to avoid bend and twist, so the designer has to consider those loads. Colin Chapman's attempts for perfection always created other problems that offset the benefits.

For a round section rod (ie solid), the strength in bending is proportional to the second moment of area Pi * D^4 / 64 where D is the outer diameter. For a hollow section, we have to subtract the bit in the middle, Pi * d^4 / 64, where d is the inner diameter. That comes out as Pi * (D^4 - d^4) / 64.

For a rectangular section bar, the strength in bending is proportional to B * D^3 / 12 (B = Breadth, D = Depth). For a rectangular hollow section, we have to subtract b * d^3 / 12, where b and d are the inner dimensions.

Thankfully, we don't need to do the calculations ourselves, because the steel manufacturers provide them in tables for stock sizes. It shouldn't be too difficult to calculate tables of mass per unit length either.

When twisting a member -- which should not be a consideration for the chassis, but applies to drive shafts and some suspension arms -- torsional strength is proportional to the sum of the second moment of area in the same plane at ninety degrees. For a round rod, you add (Pi * D^4 / 64) to (Pi * D^4 / 64), or (Pi * D^4 / 32). For a round shaft, the number comes out at Pi * (D^4 - d^4) /32.

Doh: Edited for maths correction.

Edited by Charlieman, 21 May 2012 - 19:22.

[GMACKIE]

Thanks for clearing that up, Charlieman.

[elansprint72]

snip:

... was the 59 the first formula car ( i.e built for an approved FAI, SCCA or RAC formula) to use square chassis tubes. (sic).


?

[Lee Nicolle]

And in many spaceframe chassis of the sixties and early seventies, the bottom of the chassis was square tube while the rest was round... to facilitate rivetting on the undertray.

Most Supermodifieds, midgets from the 60s to early 70s used a square base. Modified rules were 2" square main chassis. As are I believe V8 Dirt Modifieds to this day. Those chassis used round tube strengthening underneath to which the belly was attached.
These days I hope noone has rivetted any panels to their spaceframes of whatever construction as they let water in and cause the frames to rust out, sometimes fairly quickly too.
A lot of restored midgets and modifieds have had all the lower bars replaced in part or total because of this. Banned in speedway, all bolt on stuff must be attached to tags welded onto the frame.
The same for speedway sedans etc too.
Some clubbies etc worry me because of this, how rusty are those frames after 20-40 years.

[Doug Nye]

A square tube will have 4X the diameter, a round tube will have 3.14 ( ish) the diameter so square is about 25% heavier in the same wall thickness.

However they could use a thinner wall thickness with square , say 18 guage not 16 guage so the weght difference is not great if properly designed.

Forgive me for being dim, but "4X the diameter" of what? Can a square form have a diameter?

DCN

[Ray Bell]

He means four times the dimension of the square...

It had me fooled for a second or two, too!

[mariner]

Thank you Ray Bell, sorry Doug Nye!

English may be my native language but I can mangle it as well as anybody!

[Steffe Ornerdal]

Maybe a little off topic... but in the late 60's a speedway motorbike was constructed in Sweden with a frame of square tubes. A complete disaster. Some world class riders tried the bike and went right out to the fence, crashing. Was not rideable at all due the square tubes.

[BMH Comic]

Maybe a little off topic... but in the late 60's a speedway motorbike was constructed in Sweden with a frame of square tubes. A complete disaster. Some world class riders tried the bike and went right out to the fence, crashing. Was not rideable at all due the square tubes.

I get it that it may not have been rideable due to the construction, but it wont be because of the tube, may have been how the tube was used.

The product is able to be formed into a curve, it does loose some of its integrety in one direction only but not much less than round tube, the other major advantage to using it and this was recognised very early on in its use is that you can load the tube at its optimum point when you are fitting a socket or a mounting point. It can be prestressed as easily as a tube also.