Anyone install coil overs instead of using torsion bars?

Well I’m not sure about the floppy part. As a guy from the 60’s I well remember the typical body on frame convertibles from the era, or how about an Austin Healey which was a semi moncoque. Everything moving, a lot, in different directions all the time. I suspect anybody young today with no experience with these cars would get out of the car after a few miles and wonder if it was actually going to fall apart before their eyes- “No, but in time cricket - in time” says the teacher.

By any comparison the E Type was an utter paragon of stiffness - it’s still impressive. A large heavy bonnet that is virtually stationary at all times and under all conditions. Fifteen years of abusing one on and occasionally off a race track, hard frequent braking from 145 mph, tires with a grip level you need to experience to believe, with out even one flinch from the structure - ever, and this on the original engine frames on a car that was pulled from a farmers field. (looking back I’d probably reconsider using old frames like that - maybe got away with one here.)

3 Likes

Fair points, relative to its era. I was thinking of it in overall terms, where I see two long, narrow pylons not deeply attached to each other - like a man with two outstretched arms. It would be stiffer if they were joined with more triangulation in front view but obviously I defer to your direct experience and satisfaction.

(edit)
…although, weren’t Pressed Steel telling Jaguar back in the '60s that the frame lacked stiffness? Our tests confirmed that in objective terms, although I accept that the world at large is satisfied with it.

1 Like

That’s an interesting point - could that be why the main diagonals run to the area where the bonnet is attached, leaving suspension loads to find their way back to the main structure as best they can?

It’s an interesting point: a well -adjusted bonnet isn’t absolutely stationary, but it really does not move much, at all.

Hi Clive. Didn’t I send you a diagram of twist etc testing done on the tub and frame? I think so. The “weakness” (relatively) was the flex in the fire wall where the frames joined it.

Was that the period report, reproduced in one of the books? I do remember that, I think it pointed to the upper, outboard scuttle mount as lacking significant linkage to carry loads back into the main structure. That’s where increasing the stiffness of the horizontal connections between the two “pylons” would help to make the entire front end a more unified structure.

I should clarify the area I’m talking about when I suggest the stiffness is low:
It’s not about strength and durability, obviously Jaguar took care of that.
It’s not so much about twisting of the structure affecting suspension angles and hence cornering power, the driver compensates for that (especially on a smooth race track)
It’s more about the structure vibrating / shaking in response to small(ish) inputs from road events and irregularities. Jaguar would have tuned the suspension to minimise that but cars will have changed critical parts across the years. Tires and dampers are the most important, changes increasing the chance of shakes and disturbances. It may not be obvious and owners will have varying tolerance. To some it will be part of the period charm, others will not even be aware. For me it’s a fixation so I’m always looking for causes and cures.
Thanks for your patience.

Sorry, I do get a bit carried away with the data. I was trying to show that there are various factors putting significant loads into the front structure and coil springs would not be the largest of those - but they would not be insignificant.

1 Like

Hi Terry, I’m thinking you mean the large diameter coil spring that sits on the lower control arm, as opposed to the E-type style where both ends seat on the damper?
image

Coil springs are chicken-and-egg devices: stiffness is dominated by wire diameter and so is stress – but in the opposite direction. So to keep stress low you tend to want larger wire diameter, that can push you to more coils to avoid excessive stiffness. If you take that approach too far the coils clash as the suspension travels, so you go back and do another loop.
A heavy luxury car can have a particular conflict – you want a soft spring and long travel, so you make the overall spring diameter larger to accommodate the total wire length. It’s now too large to fit on the damper so you make a spring seat on the lower control arm, that introduces another conflicting parameter – stability. That’s the tendency to buckle when compressed, like the long skinny spring in a ball point pen. The control arm tilting relative to the spring axis can cause instability, driving the overall diameter to increase further. Etc.
Interesting things, springs.

2 Likes

N.B. four coils in the rear.

The fact that the ear breakage on the front has been observed suggests not all of the road force is adequately taken out by the damper and coilover.

A graph of what the component deformation is, is sort of half-useful in that what is really wanted is a graph of what is the total loading on the suspension MINUS what the coil plus damper can take care of. Obviously there is a shortfall or less headroom than desired as picture frame mounting point failure has been observed.

Mike’s additional picture frame bracing makes sense for a race car as the momentary loads on the damper will be higher than that expected from a more sedate road car driver.

The point about torsion bar moments also means the frame/tub/reaction plate between them must be permanently in a state of “anti-twist” moment by the same amount. Switching to front coilovers replaces those twisting moments with resultant vertical loading somewhere else. I presume that’d permanently stretch the bottom of the engine frame and compress the top.

The bonnet is quite a rigid structure if you look at all of the cross braced panels plus the fact that (aside from the OBL cars) it sits bolted down to the scuttle on rubber bushes in opposition to being pushed into the underpan by the bonnet opening springs.

kind regards
Marek

Thanks Marek, plenty to think about there. I’ll start with the easy one:

“N.B. four coils in the rear.”
I think you’re referring to the quantity of springs – I was actually using coil in the literal sense, meaning additional turns of wire in the spring to reduce its stiffness.

“The fact that the ear breakage on the front has been observed suggests not all of the road force is adequately taken out by the damper and coilover.”
It seems such failure has been observed on a car in original condition (no added coilover springs) when subjected to extreme road loads. As others have remarked, this could suggest the original frame does not have great headroom for increased load. The question is how much additional load a coil spring would apply.

“… what is really wanted is a graph of what is the total loading on the suspension MINUS what the coil plus damper can take care of…”
I believe my previously-posted chart offers that – copied below, it summarises the relative magnitude of loads from various sources. The user can add or subtract loads due to springs and dampers, according to area of interest. I suggest, though, that the phrase “minus what the coil plus damper can take care of” is misleading if it implies these components “contain” load without passing it from their input to output attachment points.
The spring develops force as a result of being compressed – it passes that force to the body. Depending on the spring rate and respective weights, the disturbance (motion) will be lower than it would have been without a spring to deflect, but force as a result of the bump there certainly will be. More on that below.
Likewise for the damper. The popular term “shock absorber” is a misnomer, it doesn’t do that. It develops force to restrain the spring from its natural tendency to keep bouncing in response to the initial compression. That is to say, it damps the vibration of the system. This can can require large forces, as shown in the chart below. Naturally these forces are passed to the car body, the system would not function otherwise.

“Mike’s additional picture frame bracing makes sense for a race car as the momentary loads on the damper will be higher than that expected from a more sedate road car driver.”
There we differ. I submit that vertical (bump) loads applied to the car in the wide range of potential road use are much higher than in racing. (Excluding cornering and braking, which are outside the scope of this discussion). My reasoning is simply that race circuits are relatively smooth – any bumps are there to add a dynamic challenge, not to cause breakages and failures, for obvious reasons. Public roads, in contrast, offer a vast range of potholes, craters, raised drain covers, sunk drain covers, curbs, speed bumps…all capable of causing unexpected and immediate damage to suspension and wallet. Vehicle manufacturers address this by building durability test courses with extreme features which would be untenable on a race circuit, yet the car is required to survive.


I’ve seen a car which could survive a 24 hour dynamic test around a local race circuit, including the usual selection of bumps and swells, lose an entire wheel & suspension assembly at the first pass on a durability track.
This might be a suitable point to offer my opinion on the forces in and around the front suspension, which have been a popular topic.

Starting with the spring:
The purpose of the spring is to reduce the disturbance (motion) applied to the vehicle body and occupants in response to a road input (bump).
The force developed in the spring and passed on to the body (F) depends on the spring stiffness (k) and displacement of the wheel (x - in broad terms, the height of the bump).
From F=kx, the basic definition of spring behavior: driving over a bump height x with spring stiffness k gives an upward force kx into the spring, which is passed on to the body. From the equation above –
Lower k (softer spring) reduces the force on the body
Lower x (smaller bump) reduces the force on the body.
How much the body lifts in response to that force – if at all – depends on the relative weight of the car sprung mass (body + occupants) to the unsprung mass (wheel, tire and some associated bits). That’s why we care about the sprung : unsprung mass ratio – the higher it is, the less disturbance passed to the body as the wheel is pushed upward.
So those are the factors affecting the force on the spring. It’s that simple: force into the spring from the road = force out of the spring to the body.

image

(Note: this and the damper analysis below are first-level general descriptions to illustrate the principles of system operation and adjustment. They are not complete system analyses but they do offer insight to the main factors at work).

The other forces to consider are in the damper – here we’re interested in the upward velocity of the wheel as it travels over a bump - not its displacement – because the force developed in the damper depends on the speed at which it is compressed or extended. That depends primarily on the car’s speed, wheel diameter and the bump height, as in:

The above equation, experience and common sense all tell us that hitting a bigger bump at higher speed increases the upward velocity of the wheel, compressing the damper faster. Faster compression (or extension), increases the force developed in the damper – sometimes at a dramatic rate, as in the chart below. That force is passed directly to the body – that is its purpose. Rubber bushings are frequently used to reduce noise and harshness at the damper mounts, they do not significantly affect the force levels.

image

The point about torsion bar moments also means the frame/tub/reaction plate between them must be permanently in a state of “anti-twist” moment by the same amount. Switching to front coilovers replaces those twisting moments with resultant vertical loading somewhere else. I presume that’d permanently stretch the bottom of the engine frame and compress the top."
I see that spring loads applied to the top damper mounts would put a compressive load there. I’m struggling to picture a commensurate tensile load in the lower frame member, I would be interested to see the analysis that gives that result. If your term, “permanently stretch” implies a permanent yield, ie something beyond normal elastic work cycle of a steel part, I have not seen the analysis to support that.

“The bonnet is quite a rigid structure if you look at all of the cross braced panels plus the fact that (aside from the OBL cars) it sits bolted down to the scuttle on rubber bushes in opposition to being pushed into the underpan by the bonnet opening springs”
No doubt rigid within itself, do you think it would enhance the overall stiffness of the vehicle structure? I have doubts about that.

Thanks to anyone who made it this far.

2 Likes

Hi Clive…iv read it…its a bit long winded…what are you trying to achive?..It should be quite obvious that if you remove the torsion bars and reaction plate as per the original post to replace them with coilovers then structural work is going to be needed…How much is another story…to do it safely it would need a structural engineer to calculate…Steve

1 Like

… and one sentence later, you say…

This is why it is hard to follow what is being said, because energy being used to deform either the spring or the damper isn’t going into the body.

It did occur to me that race circuits would be smoother, but I have no feel for how that would have compared to road conditions. I have been around the MIRA test facilities in an MG Midget - highly entertaining!

This is implying it only ever takes part of the load seen by the spring, but you have also said that damper loads can be greater than spring loads, so should you not also add that it acts to oppose the initial movement of the spring, hence the term “shock absorber” wouldn’t be wholly inappropriate?

kind regards
Marek

The weight of the car is currently supported by the fulcrum block on the underside of the car, balanced, between the wheel on the outside, and a spring on the inside (torsion bar). That spring is tensioned by the reaction plate attached to the cowl, but does that matter here? I think that’s irrelevant. If you replace it with a coil spring the weight of the car now hangs between the outriggers, but the load is virtually directly above what it was with the torsion bars, with the two points connected by a very stiff vertical part of the picture frame. I don’t see why the stress in the tube frames back to the tub would be any different from what it was. The majority of the cornering forces would continue to be fed into the frame by the bottom control arms. The position of the spring might just improve excess roll in the corner.

When I worked at Ford, in the '70s, it was not uncommon to have cars literally break in half on the Belgian block section of the torture track, especially cars with T-Tops. Dashboards, also would break, or come detached from the body.

I agree. That’s why I think the question of stress with an added coil spring only needs to focus on the “ears” that hold the top eye of the damper.

I’m sorry, I’m doing a horrible job of explaining my mental process! Maybe I’ll try once more then we’ll put it on hold till we have drinks in our hands?
(Or we could just ask Terry for a summary?)

I’m not seeing a contradiction in those words - deflecting one end of the spring creates a force which must be reacted at the other end (“for every force there is an equal and opposite reaction”). The force is expressed as (stiffness x travel).
Surely the energy required to deform the spring is only stored in the spring until the body or wheel moves relative to the other, releasing the energy and restoring the starting condition? That relative motion is caused by the force being transmitted from one end of the spring to the other.

But the damper load isn’t directly related to the spring load - the spring reacts to deflection, the damper reacts to the speed of that deflection. So, a spring rated at 50N/mm would generate a force of 500N on a 10mm bump, and 1000N force on a 20mm bump. If the smaller bump, though, is shaped to lift the wheel faster it would generate higher damper force than the larger bump.
The damper attempts to control all movement of the suspension, in both directions, according to the speed of that movement. I really don’t think shock absorber is an accurate term for that activity - in some situations “shock creator” might be more accurate!

Les (or anyone with a car and tape measure at hand) -

  • could you give me dimensions A and B in the attached picture please?

And, does anyone have pictures or other information showing failures in this area due to road load / overload?

Clive,

A= 3.25”…,.,…B= 2.375

Looks like that area could “easily” be strengthened by adding some L shaped material that would spread the load from the upper shock mount down the leg of the frame. Could likely do that using the existing bolt locations.

Another thought might be to move the upper mount location closer to the frame leg, if theres room, as in drill new locating holes.

1 Like

Yes, there are various possibilities depending what parts would be affected and how much a person is willing to change. The loading is vertical so I would look to add vertical section if possible - strength will increase by the square of that height, compared with a linear increase for a horizontal thickness change.

My estimate from the loads chart is that we’re only talking of around 20% load increase from a coil spring so it’s not a big challenge. I’m thinking a crack is likely to start in the corner so bridging across there would be attractive if the wishbone wasn’t already there :thinking:
Maybe feasible for people who’ve inverted the top mount for camber gain?


A plate across the top would gain vertical stiffness, better if it could be a shallow channel.
If adding vertical section is not practical, the usual solution of plating the sides is simple. It probably reduces tendency to buckle, in addition to direct strength increase. 1mm section increase should get the right level of gaiin?

Same as usual Steve - respectful and constructive exchange of knowledge and ideas.

It’s not very complex - we’ve established the load path is straightforward, the local structure likewise. I think a first year Engineering student could do it. I could even make a fair attempt using my faded textbooks to supplement my faded brain.