Chassis stiffening

[GTVeloce]

I was looking under my neighbours NC MX5 and noticed it has factory significant chassis bracing. I then noticed in the engine bay an interesting factory strut brace that not only linked the front struts but also attached at the rear of the engine bay.

GTV(6)'s don't exactly have the stiffest chassis but I rarely see much stiffening done. Some race cars have some bracing welded in the engine bay and the odd strut brace but never on a road car. Welding is a bit extreme for your average road car but I'm wondering why some simple bolt on chassis bracing isn't more popular? Surely it wouldn't be too hard to do. And what about a rear strut brace - I wonder what effect that would have?

Thoughts? My feeling is if we can stiffen the chassis we can improve handling without ruining ride quality in the way hard springs/shocks can.

[105gta]

I think you'll find that because of the chassis design And shape there is no simple way to bolt anything in without some welding/drilling etc... and most people don't want to do that so it becomes a racing mod only item. 

[Duk]

Rather than welding attachment points to the chassis, that you then bolt your bracing to, you could look at using structural adhesives to do the job.

A couple of advantages to using structural adhesives:
The load is spread over a greater area of the material. That's quite important when working with sheet metal and you are looking to, potentially, put a significant amount of force into a small area.
No heat related distortion from welding and not killing any internal rust protection for box sections.

Naturally there are some down sides to doing it that way, tho:
Cost. Structural adhesives aren't cheap and you have to use 1 that's suitable for the environment (heat, vibration, water, etc).
You'd probably have to get the area of the attachment back to clean metal and then repaint the whole thing afterwards.
Time consuming. These things have a cure time and that's probably going to be 24hours or more.

There's a good documentary on Youtube about the Lotus Elise. The extruded aluminium chassis was bonded together rather than being welded.

[julianB]

Something I have seen done a few times recently, which I really like, is the addition of some 3-5mm plate steel to the underside of the chassis rails.
In two cases, the cars, both GTV6s, had 75 power steering systems in them, which they removed due to the reduced steering effort from the added rigidity

[Colin Edwards]

The Abarth Spider features 3 heavy gauge aluminium plates bridging the chassis "rails".  The front plate doubles as the engine under-tray and tends to extend the front splitter. 
Although ribbed for stiffness, these plates clean up the underneath of the car reducing lift and drag a smidgeon. 
Bridging the chassis rails, these plates will increase torsional stiffness a bit by effectively creating a dual skin floor / sort of box section floor.  The vertical section of the chassis rail is the web and the two "floors" are the flange of the beam.     

Given the Abarth is 50% Mazda ND MX5, stiffening a roofless car this way is about the only efficient option.  Also helps lower the COG.  Fiat also went to the trouble of adding a very sharp edged and efficient NASA duct to the under-tray.  Only way to do this in heavy gauge plate it to make the duct a separate plastic moulding - very clever these Italians!

I figure the combined weight of the 3 aluminium plates would be no more than around 5kg.  These plates are fixed to the chassis rails by lots of 8mm bolts.  Probably weigh nearly as much as the plates!

[Duk]

The Abarth Spider features 3 heavy gauge aluminium plates bridging the chassis "rails".  The front plate doubles as the engine under-tray and tends to extend the front splitter. 
Although ribbed for stiffness, these plates clean up the underneath of the car reducing lift and drag a smidgeon. 
Bridging the chassis rails, these plates will increase torsional stiffness a bit by effectively creating a dual skin floor / sort of box section floor.  The vertical section of the chassis rail is the web and the two "floors" are the flange of the beam.     

Given the Abarth is 50% Mazda ND MX5, stiffening a roofless car this way is about the only efficient option.  Also helps lower the COG.  Fiat also went to the trouble of adding a very sharp edged and efficient NASA duct to the under-tray.  Only way to do this in heavy gauge plate it to make the duct a separate plastic moulding - very clever these Italians!

I figure the combined weight of the 3 aluminium plates would be no more than around 5kg.  These plates are fixed to the chassis rails by lots of 8mm bolts.  Probably weigh nearly as much as the plates!

Interesting. A photo or 3 would help a lot.
Keep in mind that if you bolt aluminium to steel and have any moisture, which you will, you will also get electrolysis and then rust.
Aluminium, even the best grades aren't as stiff as an alloy steel like 4140.
And carbon fiber is much stiffer than steel based alloys.

[Colin Edwards]

Photos may have to wait a bit.  I don't own a hoist and as low as the car is standard, I have lowered it a further 20mm!!

The aluminium plates appear to be clear coated and given the popularity of aluminium exhaust heat shields, seems manufacturers aren't concerned about electrolysis.  Decent surface preparation by Fiat / Mazda probably helps there. 

These plates don't need the modulus of steel or its alloys as they operate in tension and compression - just like a beam.  These plates resist the chassis rails moving fore and aft realitive to each other.  They will not offer much resistance to any relative vertical movement.   

I also replaced the driver and passenger side door location rubber blocks with equivalents made from delrin.   This causes the doors to be triangularly loaded, further resisting chassis twist.  Well that's the theory!!

[bonno]

Interesting topic and as such some research on the net, provides a solution to strengthening the Alfetta GTV chassis which is in the form of boxed crs316 grade sheet steel bolted to the floor. Refer to the following link on a company that manufactures kits, but unfortunately not for the Alfetta GTV. Also have included a photo of underside of Alfetta GTV identifying the chassis sections that need to be treated (no engineering for stress and torsional rigidity / backyard engineering only) .
Chassis stiffening/strengthening
https://circuitdreamer.com/product/parts/chassis/parts-chassis-frame-rails/rad-roadster-garage-star-stainless-steel-frame-rails/

[Duk]

The GTV6 had two chassis variations which was more rigid? the 1st series or the 2nd series of cars with the Alfa75   chassis

I do remember reading (probably from someone on GTV6.com) that the 75 chassis had something like twice as many spot welds (robot assembly line) as the original transaxle chassis had.

This sort of thing requires real data.

"Without data, all you have is an opinion."

Building a suitable test jig/arrangement would be required for genuinely accurate information.

Attaching it to the car would be another box of frogs.

The front is quite straight forward. Attach the torque applying arm to the chassis mounts for the lower control arms.
But if someone was looking to use coilovers, the arm should be attached to the top damper mounts.
The front test jig should have 1 pivot point that attaches to the arm, between the chassis rails and sits on the floor.

Attaching the test jig to the rear is the hard part.
Ideally the twisting force should be resisted thru the chassis rear spring seats.
But how do you attach a frame securely to a pair of chassis spring seats???
I had thought of using some large diameter pipe to replace the rear springs, but then you have to clamp the dedion's spring seats.
So I thought about using threaded/stud bar thru the damper mounts to pull the whole thing together, clamping the large diameter pipe between the top and bottom spring seats.
The 'maybe' problem that I see with that, is that the damper mounts would be under tension and during the testing could be a point that flexes a lot, potentially damaging them and maybe skewing the results.

Alternatively the rear part of the test jig could be attached to the transaxle crossmember.
But this would be less accurate and you would be putting vertical forces into parts of the chassis that are not intended to have those forces put into them and that could really skew the results.

[Duk]

A while back, I made a scale model, wire frame model of the chassis, from the firewall forward.
I made it from TIG filler rod. It included all of the major box sections and chassis rails. The firewall was simply made as an 'X', with the cross point welded together.
I never took a photo of it, because I figured it would be really hard to figure out what was going on.

The wire frame model was quite easy to distort torsionally, as these things are considered to be.

I then added some additional bracing to the 'chassis' and they consisted of:
A 'K frame' brace under the bonnet, basically the same as the I made for my 75 (still not tested..........).
Triangulated braces that that live inside the front guards and go from the base of the A pillars to tops of the box sections that the top damper mounts and top control arms attach to.
And finally, a lateral brace that in an Alfetta/GTV type body, would basically live where the radiator lives (the 75 obs has its radiator further forward).
While the whole thing was crude, each addition made the frame harder to distort.
In fact, the wire modelling changes were so effective, it has me struggling with justifying the time and cost of any other testing (in nearly complete contradiction to my previous post............).

[Colin Edwards]

Seem to recall ADR 29 introduced in the mid 1980's.  This Australian Design Rule stipulated a certain impact load rating for passenger vehicle doors.  Cars complying with this requirement usually featured an assortment of bars and beams and plates and webs etc within the door cavity.  The intention was to limit the ingress of another vehicle into the passenger compartment via the doors.  Also recall this extra metal in the door made installing big audio speakers a real pill!!

If your GTV6 has this bracing within the door cavity, it could be used to marginally stiffen the chassis.  The presence of doors in a road car creates a fairly large cavity in the chassis.   The stiffness of the Abarth Spider possibly benefits a bit by using the doors to bridge this door cavity.  An interlocking male / female joint is created when the door is closed.  A rubber block (the male bit) fitted to the chassis couples with a female receptacle fixed to the door.  During cornering when the chassis twists, load is transferred through the door structure via compression of this added interlocking joint.  I have replaced the rubber widgets with a Delrin equivalent.  The Delrin has almost zero compliance compared to the original rubber.  See attached pikkies.

The larger the door, the larger the chassis cavity and the more likely chassis twist will result.  Two door coupes would be particularly affected by this situation.  If the GTV6 doors do feature anti-intrusion technology or it can be added, fitting this interlocking joint as used on the Abarth (and possibly the MX5?) may be of some value.  Might help - cant hurt.       

[Colin Edwards]

The female bit

[GTVeloce]

Interesting Colin. I haven't checked my current 86 GTV but my old 83 GTV's all had seriously heavy doors with chunky side impact metal bars inside them. I always thought if you were making a race car its the first place I would save weight (assuming a roll cage was installed of course). The GTV doors are also very long - much longer than a 75 for example.

Interestingly, the 75 does have a section of stiffening bolted to the underside. One on either side from roughly behind the end of the torsion bar to just before the de dion.

[GTVeloce]

I've stolen a picture from Shinycar of the brace I am referring to. This just bolts to the underside of the car with two (or maybe three - I can't remember) bolts.

[Colin Edwards]

Interesting Colin. I haven't checked my current 86 GTV but my old 83 GTV's all had seriously heavy doors with chunky side impact metal bars inside them. I always thought if you were making a race car its the first place I would save weight (assuming a roll cage was installed of course). The GTV doors are also very long - much longer than a 75 for example.

Interestingly, the 75 does have a section of stiffening bolted to the underside. One on either side from roughly behind the end of the torsion bar to just before the de dion.

If we are talking race car then installing a cage for safety and increasing chassis stiffness is a given.  However if we are "tuning" a "road car" and installing a cage is off the table, other means of chassis stiffening would need to be explored.  Using the door to structurally triangulate and stiffen the door cavity in the chassis may be worth a ping especially if the door itself is relatively rigid.  As most doors (when closed) are fixed to the chassis at three points, introducing a fourth should increase dimensional stability in that area..........in theory?!?

Worth a try?

The two square section tubes bolted to the underside of my 75 certainly dont look much.  I did remove them both, jacked up one corner of the car and then attempted to re fit the tube.  The holes were misaligned by about 3mm.  Although not much to look at, these tubes must be doing something.