Detachable Roll Bar for a historic single seater

[Masser]

I hope someone may be able to guide me on this.

I have an historic single seater which currently has no roll bar. I was wondering whether it would be acceptable to have a "bolt on" roll bar, or a roll bar that can be detached in some way. I do take my own safety very seriously indeed and I would propose that the roll bar will be in compliance with the FIA specs. I am aware that the standard is to weld on a bar but there are various reasons I would really like to avoid this if possible.

Any guidance, particularly from any scrutineers, would be most grateful.

Thanks

[D-Type]

As a structural engineer, may I remind you that a bolted connection is often the weakest point in a structure. In the case of a roll bar the range of potential loadings is vast so you must be sure that the connection has sufficient strength in all six directions.

[Masser]

D-type, thank you for your reply. I had envisaged this and I would be happy to weld on base plate for the roll bar onto the chassis rail to give the bar a very solid and secure fitting point. I would also envisage fitting a detachable brace to a mounting point to the rear. At the end of the day if I am not happy that it will do the job then I will go the conventional route.

[ss_collins]

my Austro Vau has a detachable brace but the hoop is integral

[dtype38]

Interesting topic considering what's going on in the "Safety in Historics" thread... I have a suggestion for your "detachable" hoop. I don't know if this idea is to the approval of the FIA, but it's how my own main hoop is attached and I am happy to trust my own life to it. I'd also be interested in D-types opinion (if he's talking to me that is ).

The method is as follows: The bottom ends of the main hoop are simply cut open ends of tube. In my case the hoop is made of 50mm dia CDS with 2mm wall thickness. The bodywork has two holes which the open ends feed down into towards the chassis. On the chassis itself I have two 45mm dia spiggots welded and braced to the chassis pointing up at the holes. The spiggots are about 8" tall, taper to about 40mm dia at the top end, and are capped and welded closed at the top. The powder coating on the spiggots make them a close fit inside the main hoop. Connection is made via two cross bolts at 90deg to each other spaced evenly and below where the spiggot taper begins (this prevents there being any "corner" which might cause the main hoop tube to fold) The spiggot also has welded cross tubes to prevent it being "squashed" when the nuts are done up. BTW the main hoop has its cross diagonal already welded in and that drops slightly below body height to 6"-8" abouve the top of the spiggot. The gap between is where the door bar attaches. When the hoop isn't fitted I have some shaped aly plates which rivet over the holes.

If that's all double dutch let me know and I'll try and do a sketch

[D-Type]

It's difficult to visualise without a sketch. I don't really follow it but it sounds as if the whole load might be being put into the cross bolts. If so, can they tear through the tube?

[dtype38]

Hmmm, tried posting sketch but that option seems to have disappeared from my browser. Think somebody is trying to tell me this is too boring?? Not to worry, sounds like my design wouldn't be suitable for Masser's car anyway.

[D-Type]

Your connection will be fine in any sort of lateral load situation as the tube would bear directly on the spigot. However if you load it axially (e.g. pick the car up by the hoop, or drop it vertically onto the hoop) I'm not so sure. The load has to go from the tube through the bolt(s) into the spigot. So the shear capacity of the two bolts must be greater than the axial capacity of the tube. Possibly more critical is the interface of the bolt with the tube. There is a stress concentration and the tube may fail locally under the bolt, the hole elongates and the metal of the tube effectively 'tears'.
This assumes the full axial capacity of the tube can be realised. In practice it would probably buckle first, in which case the strength of the connection need only match the buckling load.
To be sure of a connection being as strong I would use a pipe flange type bolted connection (or a square plate). It goes without saying that the weld to the flange or baseplate, the thickness of the plate and the number, diameter and grade of the bolts are the main potential failure points.
This is all very theoretical of course and there is no substitute for asking the specialists.

[Masser]

I have to thank you both for time and your suggestions, they are very much appreciated. dtype38 I was anticipating a lashing given the other safety thread!!
I understand your design, your description was very good. However as there is little room in a Cooper MkVIII 500 to attach a spigot I don't think it would be appropriate. I did come across one interesting, if not the most attractive, method (http://www.race-cars.com/carsold/coo...82995549lf.htm)
D-Type I can see your point on the stress concentration around the bolts and that is why I was thinking of the flange type bolted connection. If I was to weld to the top of the chassis rail a flange with a full support to make the connection for the flange at the base of the roll hoop as secure as possible thus not having the load concentrated either in the cassis rail or in the base of the roll hoop. I would be using high grade materials and bolts.
There is no substitute for asking the specialists but I don't know who to ask!! Would it be someone like Safety Devices I wanted to get a feel for if the concept would be acceptable to participate or if I missed something that would instantaneously exclude the car from an event (circuit or hillclimb).

[Dave Brand]

Originally posted by D-Type
(e.g. pick the car up by the hoop, or drop it vertically onto the hoop) I'm not so sure.

Good point about picking up by the hoop; at a recent meeting we an incident where an historic single-seater was badly enough damaged to require a full lift. The driver requested that we didn't lift it by the roll hoop, as he didn't think it would take the weight of the car; it made the recovery much more difficult, as we had to manually lift the front of the car to get straps under it. In safety terms, however, I think such a roll hoop can be perfectly adequate - lifting imposes loads which will not be seen in a crash.

[dtype38]

My connection looks similar to the pic Masser connected to (although the other way up and with two 10mm cross bolts. I think that so long as the main hoop "bottoms" inside (or outside) the spiggot tubes then it will share the load on the bolts if dropped "on its roof". In the picture shown, I think the top of the spiggot tubes should be belled out at a shallow raduis at the top end, so as not to cause local stress and therefore folding at that point in lateral loads. As the hoop is taller than it is wide, I also think it needs a cross diagonal in the main hoop, and preferably also between the back stays.

Basically, picture to yourself the following situations:

1. You're turning into a corner and some idiot tries for the inside and clips your back wheel. It pushes your back end round, your outer wheels grip, and you barrel roll across the track. Will your hoop a) stay upright and b) hold the weight of the car hitting it sideways on the top corner? Then... after a couple of rolls you end up in a grave trap upside down. Is the line between the nose of the car and the top of your hoop (which is now sunk into the grave a few inches) still above the top of your crash helmet... or do you now have a broken neck?

2. You leave your braking too late / are outbraked / are brake tested / your brakes fail... and one of your front wheels hits the back wheel of the car in front. Your wheel climbs the one in front and you're flipped over. You land upside down with the full weight of the car on the top of the hoop. Is your hoop strong enough to take the weight of you and the car falling on it upside down? Are the back stays strong enough to stop the main hoop folding flat backwards as you slide along the track.

I think if you keep these two thoughts in your mind when you design the hoop, then you won't go too far wrong.

I would also like to mention that a very experience racer once looked at my cage and asked how many seconds a lap I thought I was losing because of the weight and high centre of gravity. I guessed a second or so. He then asked if it made me feel much more secure and confident with it fitted. I said it did. He said that he didn't think it was losing me any time at all

[D-Type]

In the picture Masser linked to, the weakest point is obviously where the bolt goes through the tubes. If it goes any where it will be there.

Under either of dtype38's scenarios the load would be the weight of the car plus a dynamic effect. Offhand I would say the effective weight would be about double.

If the bolted connection did fail, the hoop would be forced down into the socket (or in dtype38's case over the spigot) until it reached the main frame member. It would hopefully have done its job but would need replacing. With a flanged connection, if the hoop hadn't distorted it wouldn't need replacing.

Seriously though, if a hoop isn't strong enough to lift the car with then it won't be strong enough to take the weight of the car will it? The words chocolate and teapot come to mind.

I think it's time for me to bow out and leave it to them as know, i.e. the hoop makers and the scrutineers.

[Dave Brand]

Originally posted by D-Type
Seriously though, if a hoop isn't strong enough to lift the car with then it won't be strong enough to take the weight of the car will it?

Without getting bogged down in detail, yes, it could be.

[dtype38]

Originally posted by D-Type
I think it's time for me to bow out and leave it to them as know, i.e. the hoop makers and the scrutineers.

Hmmm, not overly convinced about the last bit. I was in a scutineering bay and watched the car in front of me being checked. It was a small open car with a roll hoop made out of what looked only slightly stronger than bent coat-hanger wire with one light back stay. The scrutineer merely glanced at the hoop, duely gave him a ticket and off he went. When it came to my car the scrutineer examined my full roll cage in some detail eventually spotting a small includion in a weld that hadn't been filled by the powder coating. There was a tiny (2-3mm)spot of rust just starting. I was told that this could cause serious weakening of my cage if it went any further and that I needed to go back to the paddock, clean it and seal the spot properly before I got my ticket.

It was with great effort of will that I just did as I was told rather than argue the case... particularly after what had just been passed as "safe" right in front of me by the same scrutineer.

But then one always has to treat scrutineers like customers... always right

[HiRich]

Masser,
I've only just realised that you have a MkVIII

First, I would strongly recommend a fully welded in main hoop. Of all the 8s-11s I know, only one runs without a hoop, and to be honest I think he's bonkers. At VSCC Mallory this summer, Richard Fry rolled his 8. Even with a decent hoop he was badly injured - I understand that it would have been extremely serious without a hoop.
The effect in looks is marginal. The main hoop should be the same thickness as the bulkhead tube.

Assuming that you ignore my advice, let's split the job into two parts, the hoop and the reinforcement bars (back to the engine).

For the main hoop, make sure that you lower your seat right down (some sit quite high - for many reasons it's good to make up a seat that sits down in between the two lower rails) and then (as mentioned already) make sure there is helmet clearance plus at least an inch (get a mate to rest a 2x4 on the front spring carrier and over your helmet).
Most Coopers do not have the hoop reaching all the way to the upper rail/bulkhead joint (not least because all Coopers avoided multiple joint welding). You can pull the mounts inboard an inch or so.
If you must have a detachable hoop, reverse the design that you found, so the hoop fits over the stub tubes (which to be honest you can than machines from solid and weld onto the bulkhead). If you have the long bodywork, mods will be fairlt discreet.

For the reinforcement bars, you have more choice (depending on your bodywork type). Whatever you do, the bars will probably be hinged at the bottom, and pinned to the main hoop, to facilitate engine and bodywork removal.
- The tubes will need to be much thicker than your example.
- Do not flatten the ends. Your example would just fold. Better, weld a short cross tube in near the end to accommodate a pin. Your tube drops into a U-bracket on the hoop and frame, with a nice fat (14mm?) pin. Much stronger.
The most common arrangement is a single tube back to a mount near the rear spring carrier. The damper mounting tube doesn't look right on your example, and I'd be surprised if you have a mounting available there. Alternatively, you could run two tubes back to the upper rails, but make sure they go as far back as you can. To a degree, this would depend on whether you have the long or short engine cover.

In summary, there are better solutions than this example. I'll try to post an alternative, as fitted on my brother's car. If you can make it, there should be some Coopers at Cadwell this weekend and the 750MC finale at Mallory mid-October. Drop me a PM and I can ask my brother if he knows who'll be there (he runs the 500OA website at www.500race.org).

[Masser]

D-Type, many thanks for your very useful input. I think you may be correct and it may be time to put a design to a specialist.
d-type38 and HiRich I am very aware of the need for the roll bar and would not even think about competing without one!!
I had thought that there would be a better way then the one I found. I would agree that the bracing bars do look as if they would be a bit light-weight. I am leaning towards the idea of a base flange on the top of the chassis rail and a flange and bolt on one piece roll hoop with a more cross bracing and a more substantial brace bar attached in the way HiRich described above. I feel that with this design I may avoid the possibility of the hoop failing or folding at the points where it hoop would meet a spigot.
I will not be able to make it over to Cadwell this weekend but a good friend on mine is travelling over to marshal and I will ask him to keep an eye out.

[dtype38]

Dunno if the chassis design is close or not, but you might want to check out the roll hoop on the new "ERA Historic Single Seater" kit car by Tiger Cars. It only has a single back stay, but looks substantial and also has front stays which look designed for shoulder and side impact protection. There's a good set of close up pics of it in the August edition of "Track and Race CARS" magazine, and also a few pics of it at www.tigersportscars.co.uk

PS I agree with D-type that a flange connection is better. I didn't do that on mmy main hoop because it has to slot down through holes in the bodywork and flanges would have meant very big holes. My front hoop is flange mounted to the chassis though. :-)

[HiRich]

Masser,

I've got a couple of ideas for you, which I think can give you the strength, look OK, and be removable. If you PM me with a fax number and email address, I can fax you some sketches, and email you a photo of my brother's layout (which incorporates belt mountings quite well). If you do send me a PM, would you please advise which type of engine cover you have (full length, tail cowl only), and explain why you want the bar to be removable.

For detail design, I can point you at Jules Reichman, who is building a MkIV. He's pretty good on the design front (it's fifteen years since I did it, so I've forgotten all the load calculations and so forth) and of course is knowledgeable and sympathetic to F3 designs. He could probably prepare drawings (so you can seek scrutineer approval), and it may be appropriate that he makes up the parts so that your local man need only do final finishing to mate up to your existing frame.

I've asked my brother if he knows who will be at Cadwell. If he comes back with an answer (big if), I can lead your mate to the best examples and help introductions.

[Masser]

HiRich

I have sent you a PM with the details.

Thanks

Masser