Super Seven chassis design 5

[sacem1]

I'm starting to build myself a Lotus Super Seven replica but am starting from scratch not buying a kit from any supplier, I already got hold of Ron Chapman's book for the Locost S7 but find the design too fragile for fiting a Toyota 2 liter twin cam engine. The original chassis was 1" square tubing 16 swg thick with some reinforcement struts 1n 3/4" square tubing same gauge.
I think we need to raise torsional stiffness found a web page from Australia that stated minimum torsional stiffness requiered there was 4000 Nm/degree and that the Locost design did not reached it.
Got another data from someone in NZ to add some braces to raise torsional stiffness and that can also be done.
I am willing to even build several different chassis and test them to check results but would like your help in the following points:
How can I measure the torsional stiffness of a chassis after finishing the welding but with out having assembled all the other components?
I was thinking of bracing to a solid rigid heavy welded sawhorse lets say the back of the chassis, supporting the front on a center pivot point and fixing a lever to one side and measure the deflection with dial indicators as load is applied to the "torsion lever" then simple math will do, but is this the correct reading of deflection?
The same process could be applied to different planes across the chassis and we could in that way find the deformation loads that are needed for each degree of deformation found.
As all chassis are going to have the same aluminuium sheet cover applied over I want to study bare tube chassis first then after finding the best setup or construiction found, proceed to investigate if its better to pop rivet the Al sheets (like almost all S7's are built) or try some new bonding agents that did not exist in Colin Chapmen's days and that may give better results today.
Any help would be appreciated.
SACEM1

 

[GregLocock]

Your suggested method of measuring it is very close to being right. We support the body at each spring tower and then twist between the front and the back. This means building a frame for each end of the car.

Glued construction should be better than pop riveted, partly because you aren't drilling holes in the tubes.

I'm surprised braces will help with torsionals, torsional stiffness is a bear to increase without substantial increases in mass. That is assuming that the original design is sound, it may not be.

Check out the torsional properties of thin walled tubes, and how they add when you build them into structures. I think you probably want to go to round tube (preferably) and with a larger OD (essential).









Cheers

Greg Locock

 

[JakubMech]

How about paneling in aluminium honeycomb? The cost isnt that huge and there should be a marked improvement in torsion. This is what the saleen uses and also what caterham used (uses?) on its replicas.

Although with honeycomb u will have to include brackets to glue and rivit (to hold while glueing) to the spaceframe.

Greg, have you ever come across Ti paneling? I would think this would have a slight advantage over Al paneling due to its higher specific stiffness. Or is the cost just too high? corrosion (galv)?

Jakub

 

[sacem1]

Thanks to both for your answers.

To GregLocock:
So the torsional stiffness is measured using as the pivot point the other corner and not the centerline of the vehicle? that gives more actual flex per degree of flex, any how to my advantage so I won't contest that way just want to be sure.
The glued panels really eliminate , almost, the pop rivet holes that in a complete car run to about 500 rivets all around, thats also a lot of hand work, will have to design some tipe of de-sta-cog fixtures to keep the sheeting and tubes in full contact to avoid wavy lines in the finished car.
In round tube vs square tubes you enter contested terrein, I have been myself tempted to switch to round tubes as I have had very good results in previous dune buggy construction.
As a side note: one of my built buggies has gone by 3 crashes with other cars, not me driving but its actual owner a friend, and the 3 other cars had to be towed out and the buggy went home on its own and the chassis was extremly rigid, only the front torsion tube slightly bent on one ocassion and very minor dents on the others.
However the Caterham uses square tubing, but the Robin Hood and fraser use round one.
From my teoretical calculations I get the following:
For the same weight/foot of tubing you find that the I4 in square tubing is greater by about 20%+ than in round tubes if you asume that all deflections are only on the vertical and horizontal planes (X & Y coordinates)if you do the same calcs but asume that load can come in any direction then definitly round tubes come out as winners.
The thickness increase must be kept to a minimun as we are triying to keep the dead weight as low as possible.
Could use more feedback from you guys out there.

To JakubMech: The Al honeycomb would be undoubtedly better but is a material simply not found down here in Peru and the amount requiered is so small that importing it is not cost worthy as a solution, as you might guess Ti is far more out from availiable here.
Thanks
SACEM1

 

[GregLocock]

Sorry, I didn't explain the frame very well. The back of the car is secured to baseplate via the spring towers. The front of the car has a frame built that picks up each spring tower, and has a fulcrum, central laterally, and at an unknown height relative to the car (sorry I don't know how this is determined).

The idea is to measure the pure torsional stiffness of the car, with no bending.

Ti is rarely weight effective in stiffness applications, the specific stiffness is so similar to the other good metals. Ti is great for high strength, more than high stiffness.

Dropping shear panels into a spaceframe is a good way of improving the joint efficiencies - hmm maybe that's the reason for bracing the spaceframe. Converting 'pin' joints to 'fixed' joints has got to be a good thing, especially in torsion.

Cheers

Greg Locock

 

[sacem1]

Thanks Greg, I will start my trials tomorrow and keep you posted
SACEM1

 

[GregLocock]

http://www.ces.clemson.edu/~lonny/pubs/conference/sae983054.pdf

will be very helpful for your rig

Cheers

Greg Locock

 

[HDS]

Is there an SAE standard test for torsional stiffness? There are so many ways to set up the test that numbers are meaningless unless you know the test setup.

 

[GregLocock]

I don't believe there is a public standard. In industry most people are correlating to FEA predictions, and also doing a modal analysis on the glazed body in white. Using these three techniques together you can decide whether a given setup is satisfactory.



Cheers

Greg Locock

 

[sacem1]

To HDS:

I got hold of a web page from Australia that stated that to get the approval for a single construction vehicle in the technical inspection you had to have a chassis stiffness of at least 4,000 Nm per degree of torsion but it did not stated the way they mesuared it.
I have tried to reach the same web page but didn't been able to.
In that same article they stated that the design of the Ron Champion Locost Super Seven chassis didn´t achive the requiered stiffness but they do get approved in England in their Single Vehicle Aproval Inspection which is requiered to get a circulation license there, so it seems standards are not so level all around.
What Greg states is really certain to stablish a way to mesuare and then to compare results, in his recomended paper (see his reply Jan 5) you can get the data from Winston Cup racing chassis stiffness, if they can manage 200+ MPH crashes and walk from them I would say thats the upper limit stiffness requiered.
SACEM1

 

[ProEpro]

sacem1

Crash is a strength issue. Not a stiffness issue.



ProEpro

http://www.whitelightdesign.com

Pro/E FAQ

http://www.whitelightdesign.com/servicestips.htm

 

[RKPsyche]

If you want to try something really radical, I have some articles on eliminating the torsional stress in the first place. The first one transfers torsional loads directly to the opposite end of the car through the suspension, so that the chassis never has to cope with it. The second eliminates resistance to the wheels assuming a warped orientation and should perform better than conventional suspension because of better tire performance, as well as eliminating torsional loads for the suspension. You will still need some torsional stiffness to cope with engine torque and suspension bottoming, but it won’t be nearly as critical.

None of these ideas have ever been built. But since you are willing to build several designs and test them, I figure you may be interested. Both essays are at

http://members.cox.net/rseike/index.html

 

[toolmantwo]

SACEM1
Just a silly question; will the car be used for racing?
If so:
1. Do they have standards?
2. If the chassis is tested, does this in fact weaken the chassis beyond the point where it is not safe to use?

Would using tubing that is a high strength alloy be considered? Maybe 4130 steel?

If the AL sheets are not riveted (or fewer rivets are used) will that violate the standards/rules?

I might tend to ask the aircraft people for input, just the way I operate.

Sometime soon I want to build a streetrod Ford 32 sedan. I plan to visit some people who have made several to get their input.
Good luck

 

[sacem1]

Hello toolmantwo:

I'll try to answer your questions as much as I have advanced up until now:

Is it going to be raced?

Right now there is no Clubman car competion series in Peru, the idea is to get a category going and have at least 10 or 12 cars built with this basic chassis design with different engines so some sort of locally built cars start competing in amateur racing, (we have no Pro Racing here).

A new track is being built so maybe it's just the right time to start this. So I can not know if it is going to be finally raced by someone (I'm not a race driver myself) but it is going to be built to be race worthy by me and I hope I'll get someone to race it if I start the Clubman series going.

Do they have standards?

The racing categories here have standards for other type of cars so i've started some talks with people at the Auto racing club so that we can stablish some sort of rules as engine size (maybe up to 1600cc with multivalve engines and 2liters with up to 2 valves/cylinder) no EFI and standard engines with modified carburetion and cams to keep startup costs down, things are not so definite right now.

If we test the chassis is it going to weaken or destroy it..?

The idea of finding a way to test the chassis was to determine a minimun standard for building the cars not the ultimate strenght, which will cause the failure of the chassis, if it fails before the standard then we know its not up to the minimun standars.

So if no standars are stablished up to what standars should we test the chassis?, well I have taken as a minimun the 4000 N/m per degree of flexing that I understand is requiered by the Australian Authorities to accept a chassis as this much higher than the English built Locost chasis which is able to be licenced on public roads there.

The reason of the higher standard is that here the roads for the race tracks are either paved or UNPAVED so I assume we are going to need a greater rigidity in our cars.

Have we considered 4130/4140 alloy tubing?

Indeed it would be a much better option but it is not availiable here in any size much less in thin wall square tubing, and our use does not justify importing such small quantities.

For the Al sheetting we are going the riveted plus glued way, at least on the first car.

I sure would like to be able to get in direct contact with someone who had allready built one Super Seven chassis around here but as I stated before we are going in locally uncharted waters, so thats why I have to rely on out of the country advise.

For all of you who have followed up to here I will tell you that the chassis has been built and is right now in the process of waiting the chassis tests before advancing to the sheetting step, I hope it will resist the above mentioned load so that I can advance further on.

Several other possible builders have shown interest and due to the fact that we have fuil the chassis using welding jigs for the entire car the next cars are going to be a breeze to build if this one is right. The idea is to finish the chassis, sheetting, and suspension and each one to add the motor and trans (by the way I spent more in the jigs than in the chassis but I like to work with the idea of at least 10 or 12 chassis being built)

The next obvious step would be to stablish the standars but it has been a bit of what is first the egg or the hen because they want a finished car to see and test AND THEN sit down to decide the final standars, so maybe I'll just have to jump forward and risk having to change afterwards my first car to the future standards.

Thanks all

SACEM1

 

[toolmantwo]

Thanks SACEM1,

Looks like you might enjoy being a pioneer. We have some of the go cart tracks here in my small home town (3,000). They look like a lot of fun, but I have not tried the new track or the carts yet. They are the very low and wide track wheel spacing, so they can turn on a dime.

I plan to get a frame soon for the 32 Ford. I have to check here in the USA to see how to get the car to meet the requirements for travel on public roads. I am also restoring a Ford Model T 1927 coupe to original.

I think we can get 4130 square tubing here in the USA (not confirmed). I used to have good connections for tubing.

One city near me build the Bellanca aircraft. I had the plant tour and this was great. The materials as well as the process cannot be changed or it would require re-certification to fly. If you care to check it our is t is at

http://www.pilotfriend.com/jump_to_frames/bellanca_main_frame.htm

Have agreat weekend, John

 

[sacem1]

Thanks for the link, every day we learn a little more and so the world advances a tiny bit, but I do not see miself as a pioneer but as a stuborn guy who says that we can also do what others have done and more.

By the way even though I fly commercial planes a lot with no missgivings you will not see me in any personal aircraft and even less involving in construction of one, would not let another guy pilot it if I had not tried it myself first and that I would not do.

Its great you have plans for that roadster, they are really reminisent of my old Hot Rod magazines.

SACEM1

 

[FEMdude]

sacem1,
Have you visited locostbuilders webpage at

http://www.locostbuilders.co.uk

I think this issue has been widely discussed there, especially by a fellow named "cymtriks" He has done an FE-analysis of locost chassis and has given a lot of suggestions to reinforce it further.

Isn't Australia's recommendation for torsional stiffness (by Queensland transport) 4 Nm/degree for each kg, for vehicles with mass under 1000 kg? As the weight of a locost is eg 600 kg, 2400 Nm/degree would be enough??

 

[FEMdude]

Here's a link to the FE-analysis of a locost chassis:

http://locost7.info/files/chassis/kitcaranalysis.doc

 

[sacem1]

Well thanks a lot FEMdude for your help and those valuable links.

Right now the chasis has been built but not covered with the Al sheeting because the mechanical part is still being done and I do not want to mar the Aluminum surfaces.

My main concern for the torsional stiffness was because I was planing using a larger and heavier engine than the usual locost builder uses.

I finally did I get hold of a 1981 Toyota Celica complete car with the R18 (2000 cc) motor, 5 speed gearbox running mechanically good but with the body full of dents, rust and cracked windows, it did not made me feel guilty to simply strp all the sound mechanics to fit into my chassis.

I have changed the size of the square tubing from 1" gage 16 to 1 1/4" square gage 16 added some torsional stiffeners mainly in the front suspension supporting frame and welded steel sheeting in front instead of aluminum one.

By the way the R18 engine fits perfectly, just took the radiator fan out, even the throttle linkage for the carburetor could be used and the transmission shaft I just discarded the first section and used the second one attached to the output of the gearbox without any need of cutting or machining anything.

As this is, supposedly, the first of several S7 type cars that will be built in here your info will be put to use in building better cars everytime.

Thanks a lot

SACEM1

P.D. By the way I'm giving you a star in each of your posts.

 

[Majik]

I find this thread very interesting as I'm building a super 7 as well. Cept instead of a toyota 2L 4cyl I'm going to be running a mazda 13b turbo. If you have been around race cars before then you will notice the construction of the frames on a super 7. the tubes with the aluminum rived on every like 1" is nothing new. Alot of 70's formula atlantic/ supervee cars were constructed this way. I'm building mine out of aluminum tubing and aluminum panneling in a similar construction. in general these cars employed just a riged box up front and in the back and joined them. I'm using a design based off of the race cars I've seen and not the locost book which has a couple of cobbled up areas in the frame area espically behind the driver and ahead of the drivers feet.