Torsional Stiffness

[NickS]

Does anyone have any idea as to the typical torsional stiffness of a current or fairly recent F1 monocoque?

Thanx,
Nick S.

[Monstrobolaxa]

To give you a general idea of F1 development during the time here is more or less the typical values for diferent types of chassis/frame used in F1 during the years:

- Space Frame - 1000 lb/ft - 1356 Nm

- Lotus 25 frame (kind of space frame but with aluminum plates welded to the bottom and sides of the frame - 2400lb/ft - 3254,4 Nm

- First aluminum monocoques - 4500-6000lb/ft - 6102 - 8136 Nm

- First carbon monocoque (Mclaren MP4/1) - 14500 lb/ft - 19662 Nm

- Carbon fibre monocoques (nowadays) - anywhere between 22000 and 24000 lb/ft - 29832 - 32522 Nm

Hope this helps!

The conversion factor used from lb/ft (imperial) to Nm (SI) was: 1.356

[Lukin]

Being pedantic here, you forgot the deg bit. ie Nm/deg.

Where did you get your info Monstrobolaxa? I read a paper by Andrew Deakin (who now works at Renault F1 in the chassis dept) and he estimated 20000-25000 Nm/deg for a F1 car. The paper was written in 2000 so Im sure layup techniques have improved.

Do you know what measuring technique they use? Some people seem to think that restraining three hubs (and replacing springs with rigid links) and applying load to the fourth hub is the way to go? To me, that combines bending stiffness and torsional stiffness in a weird and whacky way.

I totally disagree, and have found some papers on it too. To me, you have to restrain the rear and apply a certain load to the front left and the same load to the front right in the opposite direction, creating a moment about the vertical centreline of the car. You can measure linear deflection and use trig etc to find the stiffness.

Also, do you have any figures on longitudinal bending stiffness?

[Monstrobolaxa]

I got the numbers from 2 diferente books:

"Chariot Makers" - Steve Matchett (ex-Benetton mechanic)
"Formula 1 Technology" - Peter Wright (ex- Lotus Technical direector)


Absolutly right I did forget the deg!

About the testing method nothing is said But to me the 3 and forth hub method sounds logical....cause he mentions that the monocoque is "fixed" and is torsioned....conserning the bending stifness it really depends onthe way the torsional moment is applied...it would have to do with the way the machine applied it! I cant try contacting someone at Jordan to get the answer! About the longitudinal bending numbers I'll have to get back to you on that!

[Monstrobolaxa]

Just rechecked the numbers for the carbon fibre monocoques (nowadays) is mentioned in one of the books as "it is not unusual to here these numbers" so it means that some monocoque do appear with these numbers but usually are slightly less!

[NickS]

Thanks for the comprehensive replies, guys! Recently, I'd heard figures of about the magnitude you suggest, but also had in the back of my mind at the time of my post figures somewhat higher than these I thought I'd read somewhere - presumably they were in error. (Or possibly from Indycar/Champcar?)

If the tub were of larger frontal X-section, would it be possible to increase its stiffness with no weight penalty? Or would the larger panels be too flexible? (i.e. Would it be possible to increase stiffness but only with a weight penalty by simply making the thing bigger?) Obviously there'd be a serious aero penalty...

Thanks again.

[Monstrobolaxa]

In my opinion increasing the size of the panels would be beneficial untill a certain limit....over that limit it will lose torsional stiffness, like you said because the painels might flex. But my opinion is that the teams are fairly close to the limit....

There is the possibility of increasing stiffness but with a weight penalty....it's more or less what happens to the twin keel cars....the keel has to have extra composite material which brings extra weight.

[Lukin]

If the tub were of larger frontal X-section, would it be possible to increase its stiffness with no weight penalty? Or would the larger panels be too flexible? (i.e. Would it be possible to increase stiffness but only with a weight penalty by simply making the thing bigger?) Obviously there'd be a serious aero penalty....

Yeah that would work for stiffness. If you consider the chassis as looking like a tube of certain radius (r) and thickness (t) from the front, the second moment of area is approximately equal to I=2*pi*t*r^3. So increasing the distance from the centre to the material, the higher the stiffness.

Like you said though, the aero loss (and also the shorter allowable a-arm length) kills any stiffness advantage.

With composites, the stiffness when used in a chassis mostly comes down to how you join them together. If you have ever got a chunk of aliminium honeycomb and tried to twist it, you will see even with big lengths, they are still bloody stiff. Its more important to work on bonding procedures to increase joint stiffness and reduce weight.

[Guest]

and don´t forget shape is a main deciding factor for stiffness.If you got big flat panels at right angles this is not very good for stiffness ,conical ,triangulated or sherical /organic shapes will make a stiffer part as well as having no holes in the structure(Like access holes for maintainance ,Cockpit opening ,removable rear bulkhead etc...

But of course the big factor is cross section .So the weakest link in thechassis stiffness will always be in the mounting of the Engine to the chassis or the Bellhousing engine/Gearbox interface where you get spikes of decreasing torsional stiffness as at these points the loads are not evenly spread ,compared to a homogenious structure like the tub or the engine itself.Installation stiffness can be very tricky to achieve.
Wonder why nobody tried a monoblock engine tranny yet.

[Lukin]

To give you a general idea of F1 development during the time here is more or less the typical values for diferent types of chassis/frame used in F1 during the years:

- Space Frame - 1000 lb/ft - 1356 Nm

- Lotus 25 frame (kind of space frame but with aluminum plates welded to the bottom and sides of the frame - 2400lb/ft - 3254,4 Nm

- First aluminum monocoques - 4500-6000lb/ft - 6102 - 8136 Nm

- First carbon monocoque (Mclaren MP4/1) - 14500 lb/ft - 19662 Nm

- Carbon fibre monocoques (nowadays) - anywhere between 22000 and 24000 lb/ft - 29832 - 32522 Nm

Is there any chance you had the approximate weight for each style of chassis?

Also, there is a pic of an experimental Ferrari tub http://www.alfapower.nu/gallery/album58/abc that apparently passed the FIA impact regulations and weighs 31.6 kg. It's pretty amazing! Apparently the stiffness wasnt sufficient to race it.

[Monstrobolaxa]

I can only find values for an actual monocoque....and usually it weighs between 35 and 40 kg......can't find any info on the other kinds! But I've only looked into one book!....when I have the time....maybe tomorrow I'll look into others.

[Lukin]

5 minutes for a reply, thats not too bad!!!

Cheers dude, if you cant find anything dont stress about it, I will just make them up. I mean gestimate based on my engineering knowledge.... or something like that. Im past caring about this assignment anyway.

Thanks

[Lukin]

I got the Peter Wright book of the F2000 today.

The monocoque weighs 44 kg and has a torsional stiffness of 2497 kgm/deg or 24496 Nm/deg.

It's a bit late, but hope it helps.

PS This book is great!

[RacingManiac]

For measuring torsional rigidity of our FSAE car, we did the following:

Replace all shocks with solid links,

Replace rear wheels with square wheel,

Put the front of the car on a point about the front axle around centerline of the car(effectively now the car sits on 3 point, 2 rear and 1 sharp point in the middle.

Apply load to one front wheel

Use dial gage at the chassis side to measure deflection

[Lukin]

That seems a little strange. The first part of your measurement technique seems fine. Restraining the rear wheels yeah, but putting the load on only one wheel?

Is the other wheel restrained or free to move?

Is the single mounting for the front removed once the load is applied?

Have you got a pic or a schematic of the process?

Cheers