Is Round Tube stronger than Square?

Is round tube really that much stronger than square for use on a chassis?

Or is it just clever marketing?

A trip to the Stoneleigh Kit Car Show prompted me to do more research.

I knew the answer was not going to be simple – I was right. As soon as I made one conclusion, another argument raised its head.

First of all, lets get the basics out of the way.

  • 1″ square tube bends less than 1″ round tube.
  • 1″ round tube is lighter than 1″ square tube
  • 1″ round tube is less prone to twisting

With square tubes, rectangular sections and I beams you hear terms like “force oriented the strong way” or “bending it the hard way”. When the direction of force is controlled or known, for a given span square tubing is considerably stronger. However, when the direction of the force is unknown or uncontrolled round tubing is the better bet.

If you built a chassis from 1″ round tube and another from 1″ square; providing the triangulation is good; the square one would be much stronger but also a lot heavier.

What would happen if you used a larger cross section of round tubing that was the same weight as the square tubing?

This is where my confusion started and the maths started. All because several professional race car builders had opposing views.

Luckily, a lot of people had already done the majority of maths for me, putting the results in nice tables.

see here:

Square Tube

An Except from here :

Size Thickness Mass/Metre Second Moment of Area Torsional Constants
Inertia Modulus
B T M/m I J C
mm x mm mm kg/m cm4 cm4 cm3
20 2 1.05 0.692 1.21 1.06
20 2.5 1.25 0.766 1.39 1.19
25 2 1.36 1.48 2.53 1.8
25 2.5 1.64 1.69 2.97 2.07
25 3 1.89 1.84 3.33 2.27

For my Haynes Roadster / Locost style chassis I have used 25×25 – 2.5mm thick ERW Square tubing. This choice was not scientific, based more on the fact that the local steel stockist had some at a bargain price. I bought 15 metres and have roughly 2 metres left. Therefore, I have 21.32Kg (13metres x 1.64Kg) of square box in my chassis.

My tube weighs 1.64Kg per Metre and has a ‘Second Moment of Area’ of 1.69cm.

Second Moment of Area is a measure of the resistance to flex and distortion. Larger values of second moment cause smaller values of stress and deflection.

The Torsional Constants are a measure of the resistance to twisting. Larger values mean less twisting.

To get the same Second Moment of Area from similar sized round tube, I would need 26.9Ø x- 3.2mm thick tubing, weighing 1.87Kg per metre (24.31Kg total). For a well triangulated chassis, with low capacity for torsional twisting, square tube looks like the better choice (read on). For a well triangulated chassis, torsional twisting is less of a critical design factor, but for a un-triangulated one the additional torsional strength of the round tube could give advantages in certain applications.

However, this is just one (poor) example, because round wins for both Second Moment of Area and Torsional twisting for the same weight. The round tube is always going to have have a larger width, but it is stronger.

For instance, I could have used 33,7Ø x- 2mm thick tubing. Although 9mm wider, it spanks the bottom of my square box for Second Moment of Area and Torsional twisting. This chassis would weigh 20.28 Kg (1.56 x 13 metres). A saving over a kilogram to boot. I could potentially replace a few tubes with carefully constructed ‘shear panels’ and save even more. Shame I didn’t do these sums before I welded up my chassis…….

The basic Haynes Roadster / Locost Chassis could be improved in terms of triangulation with little or no weight added. I recon, with the knowledge I now have, I could build one using square tube that is both lighter and stronger. Using round tube, I could build one that is much, much stronger. Having never raced, I’m not sure how much chassis flex is an issue for these cars. My FEA guru has emigrated and I’m not about to learn how to do it myself, so I’ll just have to make do with knowing mine is stronger than most.

Round Tube

An except from:

Round Tube

Outside Dia Thickness Mass/Metre Second Moment Of Area Torsional Constants
D t M I J C
mm mm kg/m cm4 cm4 cm3
21,3 2 0,952 0,571 1,14 1,07
21,3 2,5 1,16 0,664 1,33 1,25
21,3 3 1,35 0,741 1,48 1,39
26,9 2 1,23 1,22 2,44 1,81
26,9 2,5 1,5 1,44 2,88 2,14
26,9 3 1,77 1,63 3,27 2,42
26,9 3,2 1,87 1,7 3,41 2,54
33,7 2 1,56 2,51 5,02 2,98

So why choose Round Tubing over Square?

The answer is round tube has a higher resistance to both flex and torsional twisting than square for a given weight.

NB. Use ERW tubing as it is much, much stronger than CHS. If finances permit use CDS which is even stronger still.

In the examples above, it is possible to see that you can build a chassis out of round tube that is both lighter and stronger than a square box section one.

If you have a round hole, putting the maximum size round tube through it will be stronger than its’ square counterpart. However, if you have a square hole use square tube.

You’ve gotta admit, round tube chassis do look cool!

The problems start when you have to ‘fish mouth’ the end of every tube on the chassis to get them to fit. Forget the band saw, you really need specialist tube cutting equipment (holesaw notcher) or a file and plenty of time and patience.

With a small vehicle the bulk of round tubes is a problem. As the size of vehicle increases packaging issue reduce whilst the weight and strength gains grow exponentially. In other words, if you have a large, heavy car or your chassis is modeled in CAD with Finite Element Analysis (FEA) analysis; round tube is definitely worth considering.

So why choose Square Tubing over Round?

This is where chassis builders seem to disagree. The argument isn’t over the strength of round tube for a given weight, the problems seem to lie in the fabrication.

One argument is fabrication of ‘Shear Panels’. With box section frames, the gaps between chassis rails are clearly defined with square, flat surfaces. Filing the gap between rails; such as for a bulkhead; is easy with box section. With round tube, it becomes much harder and time consuming. A Shear Panel is a flat section that has a small return on all edges; typically 5mm or just enough to give a firm location and enough to weld to. The shear panel needs to be the same material as the tube frame and needs to be seam welded around all edges, with no gaps. No pop rivets here! When done correctly, this can look as impressive as a tube chassis.

A chassis using shear panels can be extremely stiff. These panels add a whole new dimension to providing torsional stiffness in a frame. You use a significantly smaller diameter square tube but fill all the open areas of the frame with steel sheets. The result is a frame that weighs the same as it’s round tube equivalent but has no weak spots. Plus, there is plenty of room for fitting engines, exhausts etc.

This is basically what NASCAR racers do. With this method, calculating frame strength to a finite level becomes even more complicated than using round tube. Both CAD and Finite Element Analysis (FEA) software is vital.

Shear Panel

The second answer is time and skill. Building a round tube chassis definitely needs a special JIG and tools. Using tools such as an angle gauge and even a tape measure become all that more tricky with round tubing. With square, clamping bits together for welding etc is an order of magnitude simpler and quicker. Filing a square end on box section doesn’t take much practice. Getting a perfect ‘fish mouth’ requires years of practice and a fair amount of judgement; failing that a substantial investment in special tooling.

The third argument seems to be with the bulk of a round tube chassis. Packaging of components such as engines and gearboxes, routing of exhausts all become much more challenging. There simply isn’t as much room for all your parts. In order to get things to fit, tubes might be omitted and tube radii reduced etc. If the designer is forced into this situation, torsional strength can actually be lost. Basically, a lot more planning and calculations are required.

The final argument is cost. CDS round tubing is a couple times the price of square tubing and the wastage during manufacture is significantly higher. A good proportion of what you buy, will end up on the floor as filings and off cuts.

I’m just a little bit annoyed that I built the chassis first then learned why I built it like I did. I built it correctly, but I didn’t really know why.


Haynes Roadster Suspension Brackets

I'm trying to make these brackets as accurate as possible. But 7 hours a piece? Come On! Some days, you sit there with your morning brew, thinking 'I'll just knock them up, then s'afternoon it'll be rolling'. This is one of those jobs that takes way, way longer than expected. I'm ...

Haynes Push Rod Suspension

The rear suspension features rising spring rate, pushrod suspension. It's all mocked up here with dummy push rods etc. Read more here. The idea being, the more the car rolls, the higher the spring rate seen by the wheel. It's what the formula 1 boys do. However, here are my problems: ...

MT75 4x4 Haynes Roadster

I've had a trial fitment of the MT75 Ford Cosworth 4x4 gearbox and it was a little tighter than I hoped. The gear selector mechanism; which will be removed and replaced; fouled the tunnel but this was a known issue. More importantly, the gearbox's transfer casing is a little close ...

Suspension Mounts

I've done the mounts for the front suspension A-arms and the Sierra Front 4x4 mounts. I'm not the first to mount a 4x4 front diff into a Haynes Roadster and at the speed I work at it's not surprised. It's not just a case of welding on some brackets there ...

Starting Again - Cutting It Up

It was decided that, although individual measurements often had small errors compared to the drawings, all these minor errors added up to huge errors. There were 27mm differences from one side of the chassis to the other and it had over 30mm of twist. Our goal was to have less ...

Plan of Attack

Having built and modified several cars in the past, with this build I knew it was vital that I laid out exactly what I wanted to do before I even picked up a saw or file. I’d had projects that started with live axles, swapped for IRS then ...

Chassis Type

Having quickly established that most pairs of measurements were different, the next task was to check all measurements against those published. In the process, establish exactly what type of chassis I´d purchased. The fact that measurements were different to those published would not be of major concern; as no ...

Chassis Collection

Having debated building a 'super seven' style car for decades, when a chassis came up on ebay with a reserve of only £0.99, temptation and curiosity took hold. The auction finished with only 3 bids at £56.55. Surely this was a bargain??!? A 550mile round trip would answer that. The unfortunate answer ...

Translate »

Web Design by Go Web Solutions