If you’re a fan of unit-bodied Ford or Mercury cars of the ’60s or ’70s — Mustang, Cougar, Falcon, Comet, Fairlane, Montego/Cyclone, or Maverick — you’ve probably heard of a structural component called a torque box. (Other automakers used them too; they were by no means a Ford exclusive.) Depending on the model, year, and body style, these cars may have originally come with two, three, or four torque boxes, and some owners retrofit a couple the factory didn’t supply. I’ve noticed that a lot of people seem to have a profoundly mistaken idea of what torque boxes are and what they do. Let’s set the record straight.
One of the conundrums of engineering a complex structure, like a car body or suspension, is that making the structure stiff, solid, and rigidly connected can be both good and bad. A rigid structure can be great for handling and for preventing squeaks, creaks, and rattles, but it’s also a great transmitter of noise, vibration, and harshness (NVH). People who try to upgrade their car’s suspension with firmer bushings often find this out the hard way: The stiffer bushings might make for more precise steering and handling, but they’ll also make you painfully conscious of every crack and bump in the pavement. Unless you’re designing a pure race car, that gets old very quickly.
If you want to keep NVH to reasonable levels, you need compliance, meaning some part of the structure that will “give” a little to absorb the energy of the vibration rather than passing it along. Automakers have whole squads of engineers (today armed with sophisticated computer modeling tools) to find the right balance between stiffness and compliance.
What does this have to do with the Ford and Mercury models I mentioned? All of those cars are unit bodies, which means they don’t have separate frames. They may have some rails and crossmembers that LOOK like a frame, but those members are welded to the body, making for a relatively light but strong structure. The problem, like I said above, is that a welded unitized structure can transmit a lot of NVH from the engine and suspension right into the cabin.
Some unit-body cars deal with that problem by using subframes, partial frames that carry the engine and/or suspension and are attached to the body using rubber mounts to soak up NVH. However, that’s more expensive, and it weighs more, so it’s not always suitable for smaller, cheaper cars, like the old Ford Falcon or the early Mustang.
The answer Ford came up with in the late 1950s was torque boxes. Ford first used these on the front of the unit-body 1958–1960 Thunderbird.
Torque boxes were included in back on the early Falcon and Comet, and at all four corners on the midsize Fairlane. The Mustang and Cougar always had rear torque boxes, and convertibles always had them in the front. Early hardtops and fastbacks didn’t have front torque boxes until 1967, when they got one on the driver’s side. Starting in 1968, the Mustang and Cougar had front torque boxes on both sides, regardless of body style.
A torque box is a metal box that connects two longitudinal rails. (While I’m talking here about unit bodies, perimeter frames also use torque boxes.) Many people seem to assume that the torque box is some kind of structural brace that more rigidly connects the two rails, which is really the opposite of what it actually does.
Above is a reproduction front torque box for an early Mustang. Torque boxes have to be welded into place to work — they aren’t bolt-on pieces — but it doesn’t really look very rigid at all, does it? Even if you properly install it, these thin sections of steel aren’t going to do much to reinforce the box-section rails they’re attached to. However, that’s not their job.
Like a rubber bushing, a torque box provides compliance. It’s torsionally flexible, so if you apply force to it, it will tend to twist. If you weld the torque box to two sturdy steel rails, it provides a slightly flexible connection between those rails.
Why would you want to do that? Ford engineer Forrest K. Poling, executive engineer of the Ford Chassis Design Division, explained it like this in a January 1962 presentation to the Society of Automotive Engineers (SAE) about the then-new intermediate Ford Fairlane:
As the front wheels move across uneven surfaces, the upward suspension reaction forces tend to lift the front rails. However, the front rails are welded to their torque boxes, and the tendency is for the boxes to twist or rotate rather than transmit these forces back through the vehicle. … The gauge and links of the torque box itself can be so varied as to obtain the degree of compliance that is most compatible with the balance of the structure.
Here’s the slide Poling used to illustrate that part of his SAE presentation:
In this diagram, the front torque box is located at the zig-zag section where the front rail connects to the sill.
Poling’s slide emphasized the bending of the front rails, but all these cars also had rear torque boxes, which connected the rear rails with the sills. The rear boxes were in a different place and shaped differently, but they worked the same way.
If all this is new information for you, you may be scratching your head saying, “Wait, but if the torque boxes are actually flexible, how is it that they make the car stiffer?” Here’s the other half of the conundrum I mentioned above: A structure that allows some compliance — the right amount, in the right places — will FEEL more solid than an unyielding one because it keeps more of the bending, twisting, and NVH away from the occupants. Think of it like sitting in a soundproof room: If you can’t hear the construction work and screaming kids outside, you may think you’re in a very quiet neighborhood.
Now, before the Mustang first launched in 1964, Ford engineers at the press introduction apparently DID claim that they hadn’t included front torque boxes on the Mustang hardtop because they would make the closed body “too stiff.” Some of the buff book editors dutifully repeated that, and it’s since become part of Mustang lore, even though it made no sense.
Misleading press statements not withstanding, the most likely reason Ford didn’t include two front torque boxes on the Mustang or Cougar prior to 1968 was cost. Installing torque boxes on the production line isn’t nearly as much work as retrofitting them in your garage, but there is additional labor cost involved, and automakers hate spending money if they don’t absolutely have to. (That was also why the early Mustang didn’t have standard disc brakes and 170/3-speed cars didn’t have a synchronized low gear — it wasn’t because they didn’t need them!)
Two years before the Mustang arrived, Ford was so proud of the torque boxes in the new Fairlane that they emphasized them in the Fairlane brochure. The text in the right column of the above spread says:
“Torque Boxes” (shown in red) are the magic ingredient in Fairlane’s matchless ride. Mounted at the four corners of the underbody, these boxlike structures are strategically placed to intercept road noise, vibration and ride harshness transferred from road to wheels to car. By torsion (twisting) action … very slight, but enough … the torque boxes effectively absorb these annoyances before they can reach the passenger compartment. You’ll never be able to see them work, but you’ll marvel at the work they do every mile you ride in a Fairlane!
So, if you didn’t know before, you know now: Torque boxes add compliance, not stiffness — and that’s a good thing.
NOTE: I’ll discuss the structural engineering of the Falcon, Fairlane, and other smaller U.S. Ford unit-body cars in much greater detail in an upcoming post.
Thank You for the very interesting and informative article! However, that prompts me to ask a question. Other unibody cars had a fair amount clearance between the front wheel wells and the engine, but the Falcon and its successors, the Mustang, Maverick, Cougar and their derivatives had those huge shock towers that really limited engine compartment space. Why were the shock towers so intrusive on those cars? Was it because the Falcon and its successors didn’t use a front subframe assembly, like the GM X-Body cars (Nova, et al) and Chrysler A-Bodies used? Inquiring minds want to know!
Stay tuned.
It’s because the coil spring and shock absorber is mounted above the upper control arm, where GM would typically place the spring/shock between the lower control arm and frame rail. Chrysler didn’t use coil springs at all in their passenger cars in this period, but instead torsion bars that twist via the lower control arm. The shock absorbers also mounted to the lower control arm as well so there’s much more room in their engine compartments
AMCs actually used a similar layout to the Fords with the spring mounted above the upper control arm (of sorts) necessitating large spring towers, only difference was prior to 1970 the shock was mounted to the lower control arm (Ala GM/chrysler) and rather than having an upper ball joint used trunnions
Do I take it those screws in the ‘Stang rear section are so that the model may be disassembled?
I cannot believe they were designed to shake themselves loose on rough roads like a Land Rover or some Class 8 truck frames!
Those screws are temporary. The car in that picture is undergoing very invasive metalwork; I’d say it’s a convertible getting new inner rockers (which were wider than the rockers in coupes and fastbacks).
The car in the photo (of which we’ll see more in an upcoming post) is indeed a 1965 Mustang convertible, which was in the process of getting an extremely extensive down-to-bare-metal structural restoration.
I’m a Ford guy who has never pieced together a good understanding of “torque boxes”–and this helped terrifically. Thanks much for a fine writeup and the useful images.
(I’m always hearing of old cars for sale that mention torque box rust.)
For the cars without front torque boxes, does all that front frame member flexing just go into the floorpan?
Also, I’m not very connected to these Mustangs, but reading around it does seem that pretty much everyone says that adding these makes the car much stiffer. Is it correct to say instead it only makes the car seem stiffer? With the forces distributed better to the side rails and absorbed my the box itself?
Yes, and sort of. The front rails are welded to the the floorpan, so without front torque boxes, any front-end flex is transmitted into the floor and the central body structure. With front torque boxes, most of the flex is taken between the torque boxes and the sills, which reduces the magnitude of the bending and torsional forces applied to the floorpan. This is why I used the analogy of soundproofing: Inside a sound booth, it really is quieter, but it’s because outside sounds are being absorbed before they reach the interior.
Thanks. Very interesting. So there really is less flex transmitted to the occupants, but the front frame members are flexing just as much with or without the torque boxes.
I noticed too that vendors sell “heavy duty” torque boxes made from heavier gauge steel or some stronger geometry. I’d guess this would be analogous to the solid suspension bushings you mention. Actually reducing chassis flex (maybe?) but not absorbing the flex transmitted to the cabin
The main (good) reason I could see for heavier-gauge torque boxes would be for a car that was going to be restomodded with wider, stickier modern tires and aftermarket chassis modifications that would put greater loads on the structure than anything the manufacturer originally envisioned. In that case, heavy-duty torque boxes might be appropriate to ensure that they have an appropriate degree of compliance for the loads they’re expected to absorb; as Forrest Poling said, you can do that by varying the gauge and links of the torque boxes. I’m a little leery of being able to do that satisfactorily with latter-day aftermarket add-ons in the absence of actual structural analysis, though. Also, because so many people have this backwards idea that the torque boxes are supposed to reinforce the structure, they may be easily tempted by the mystique of having extra-strong heavy-duty everything and end up defeating the purpose.
I agree with your leery-ness. I think most likely it’s people trying to put on “the best” without really understanding what they’re doing. Like those that put on the biggest carbs, the strongest ignition coils, etc when there is nothing calling for rhat
When the full sized 1965 Ford was introduced, I recall reading an article about them in either Popular Science or Popular Mechanics.
They emphasized the suspension and the flexibility in its design, showing a graphic depiction of the front and rear sections taking the majority of the rise and fall actions of the road, with the passengers in the middle not being as affected.
After reading this piece, I suspect implementing the torque boxes would explain this principle.
Ouch. Whoever lifted that 58-60 Thunderbird they put the foot plate in the wrong location. Right on the fuel line and weld seam. No, no, no with my Cougar and Mustang looking the same one has to pay careful attention when lifting.
As for a Mustang torque box you can see why so many get rusted out in the Midwest and need to be replaced.
Aaron, this is a wonderful essay explaining a part of automobile engineering with which I have been heretofore unfamiliar. Thanks.
Now I finally understand what the heck torque boxes are for. Thanks, Aaron.
Looking at their implementation, I wonder what kind of reinforcement was required for the Mustang used in Bullitt?
My guess is the torque boxes would not have survived too many of the jumps.
Fantastic article, it is great to learn this stuff, many thanks.
I’m wondering how these compare with Chrysler unit bodies, especially convertibles and Hemi cars, I have heard some of the the reinforcements used on those described as torque boxes.