X-Frame Stiffening

[CTX-SLPR]

Howdy,

I know that keeping the X is not a popular option, but I got what I've got and enough money for a custom perimeter conversion frame is not on the list.

I've been reading up, searching, thinking, etc... about how to stiffen the car up without having to built a complete custom frame. I was reading the thread about the Tunnel/Backbone (https://www.pro-touring.com/showthread.php?93236-Trans-tunnel-backbone-frame&highlight=tunnel+frame)frame on an early Tempest and that got me really thinking about how to potentially impliment a lot of the suggestions though with a full boxed frame already I'm not sure how much I need to apply to something like the underdash bar and backwards.
Similarly I read the Street Rodder How to Stiffen you Chevy Frame (http://www.streetrodderweb.com/tech/1002sr_how_to_stiffen_your_chevy_frame/viewall.html) and while that seems like a good idea to reinforce the center of the X, I'm not sure how much it does for the overall stiffness of the car.
The idea I've come up with (and might draw up some evening over my long weekend) is to add a set of outer rails under the door sills since my inner rockers are rusted in a few places as are several of the underfloor braces. I'd tie the outer rails, that would hide where the inner rockers used to be, into the center backbone of the car with heavy 'U' channels to route the exhaust through on the passengers side. I've also been thinking of how to tie the firewall (and any under dash bar) to the upper spring pockets. Not sure if I should try and start the bars near the center of the cowl and then splay them out around the top end of the engine and downpipe on the passengers side to more closely follow the fenders then drop down onto the spring pockets or start out wide and just go straight down more or less. While I'd like to route a "Monte Carlo Bar" across the middle, unless I dramatically free up room with a new upper plenum under the hood, there is no space. Additionally there is a truely massive crossmember that runs between the spring pockets under the engine that I doubt a piece (or pieces) of tubing arching over the top is going to help to keep it from "squeezing" in.
For the rear of the car, there is a large stamped 'K' brace behind the rear seat that ties the parcel shelf to the inner wheel wells and the center of the floor pan at the rear axle kickup. There is also a crossmember, tubular welded to the top of the frame kickups, under the rear axle kickup and I put an angle iron cross member in front of the gas tank to hold the B-body tank in place and the rear stamped channel crossmember at the rear of the frame. Since this is a pillarless hardtop, and I am not willing to put a cage in that would add pillars, what can I do to stiffen the "hinge point" off of the firewall or any under dash bar? Run a pair of tubes down under the center console to approximately the center of the X for bracing? Box the A-pillars?

Must like with the Tempest, I'm after hiding all of this inside the existing car structure so it looks like a nice restromod on the inside vs. a birdcage protouring car. Opinions are certainly welcome.

Thanks,

[exwestracer]

That car was definitely built to use the body and frame structure TOGETHER for stiffness. That's a double edged sword... If you go to solid body mounts, you increase stiffness, but hard driving may show up in body seams moving around, doors not closing properly, etc. Obviously, the areas where the body is mounted to the frame will be the places to anchor your structural improvements.

The door posts are plenty strong as is, they just need to be tied together better front to rear and side to side. Your rocker channel idea should help, and the taller you can make them the better (take that to it's theoretical conclusion and you have the doors welded shut...).

[CTX-SLPR]

I think I'm following Ray. Unlike the unibody Tempest, the body on frame Riviera need to be treated like the body and the frame are seperate units. Increasing the frame stiffness with things like the rocker bars (the rockers go up quite a bit to the door sills so I could make them something like 3-4in tall without really exiting the original envelope) and thier cross links to tie the frame together. Likewise I can stiffen the body up as if it was it's own unit but still keep the compliant mounts between them.

Reading the safety thread and your comments about tieing the doors together front to back and side to side made me think of adding something like a modern day door bar through the inside of the door to help safety wise but also help brace that opening up some. Because the door skin actually bolts onto the door frame and you get into the guts of the door from the outside this wouldn't be hard to impliment and with some careful design, I would be able to weld it in place like it should be and not interfere with the window mechanisms.

Would an underdash bar still be a good idea in this case to tie the fronts of the door pillars together?

Is the firewall still a natural hinge point that would need bracing to the frame or possibly to the core support or its mounts?

Thanks again for your help,

[LeighP]

The car will move around....unless you stiffen it up. I'd recommend adding side rails, repair the sill inners and tie the side rails to the sills...maybe upper and lower longitudinal tubing, tied together with short tube sections (ladder style) - light and stiff. Solid body mounts, and stiffen the body with a cage structure.

Many years ago, I owned a Canadian Pontiac - 63 Laurentian - basically a 63 Chevy with a Pontiac skin. Previous owner had a 396BBC in the car ,with one head chained to the LH chassis rail to stop breaking engine mounts. He put a noticeable twist into the chassis doing that.....so these cars are very easy to bend up.

[exwestracer]

The firewall is typically an excellent torsional stiffener. How good it is depends on how far away the "shear plane" is from the door pillars, how much angle the firewall has, and how big a hole it has for the heater box. How well that stiffness translates back to the CHASSIS takes us back to the original point about how tightly the two are tied together.

In general, the key to torsional stiffness in any chassis is HEIGHT. And there's the problem...height usually means the stuff we are adding gets in the way of the squishy thing behind the wheel. Regarding CTXs comments on the door structure, it's all well and good UNTIL you get to that single point (the latch) where the door ties the front pillars to the back. That's why the door tends to be largely ignored... How about a double (high and low) latch system that firmly locks the striker pins in place?

All of this involves a lot of thought, or we end up adding a bunch of weight for very little benefit. I've worked on a number of aluminum monocoque cars over the years, and I'm always amazed a just HOW stiff those chassis are for the amount of material weight that goes into them. We can learn a lot by studying those structures; remember a monocoque car has NO tube frame, so the simple sheet structures have to carry all the loads. But again...how many people want to cut the door in half and deal with a 12" tall rocker sill, as opposed to putting a full cage in it?

By way of a more specific answer to CTXs question, I would add the rocker tubes in the frame, making them absolutely as TALL as possible, then "fish plate" the body rockers to the frame (bolted if you still want to separate them) at the front and rear door posts and ahead of the rear fender openings. From the spring mounts out to the bumpers you really won't gain much, as the spring perches are where the twist load is fed into the structure.

The cross bar between the A pillars will help if the firewall is a good distance ahead of that area. I'd make it out of the largest diameter tube you can possibly fit behind the dash. It can be very thin wall to save weight. Think about welding a driveshaft across there and you'll get the idea... One way to improve the stiffness of the floor is to fab the seat risers as boxes (out of thin gauge sheet) and run them from rocker to tunnel. If you're keeping the rear seat, build a sheet metal riser that goes WAY up and run a thin foam seat bottom. Weld in a thin plate to seal off the rear seat back bulkhead. None of this has to be very heavy, and dimpled holes can reduce the weight further at little cost to the structure.

[CTX-SLPR]

The firewall is typically an excellent torsional stiffener. How good it is depends on how far away the "shear plane" is from the door pillars, how much angle the firewall has, and how big a hole it has for the heater box. How well that stiffness translates back to the CHASSIS takes us back to the original point about how tightly the two are tied together.

In general, the key to torsional stiffness in any chassis is HEIGHT. And there's the problem...height usually means the stuff we are adding gets in the way of the squishy thing behind the wheel. Regarding CTXs comments on the door structure, it's all well and good UNTIL you get to that single point (the latch) where the door ties the front pillars to the back. That's why the door tends to be largely ignored... How about a double (high and low) latch system that firmly locks the striker pins in place?Interesting idea, it does follow that there are 2 hinges which are fairly solid connections on the front of the door but only a single latch and a squishing perimeter of rubber elsewise. Also modern door bars seem to be attaching low on the doors rather than to the rear latch point though they have a "pin and catch pocket" vs. a latch. Am I to take it from your comments that the door bar is less of an improvement than simply tieing the door to the door frame better? Definately a doable improvement since there is nothing in the rear jam to interfere with another attach point. I'm still thinking adding a bar might be worth it for safety reasons but the door structure that everything is mounted too is a single piece of stamped steel already.


All of this involves a lot of thought, or we end up adding a bunch of weight for very little benefit. I've worked on a number of aluminum monocoque cars over the years, and I'm always amazed a just HOW stiff those chassis are for the amount of material weight that goes into them. We can learn a lot by studying those structures; remember a monocoque car has NO tube frame, so the simple sheet structures have to carry all the loads. But again...how many people want to cut the door in half and deal with a 12" tall rocker sill, as opposed to putting a full cage in it?

By way of a more specific answer to CTXs question, I would add the rocker tubes in the frame, making them absolutely as TALL as possible, then "fish plate" the body rockers to the frame (bolted if you still want to separate them) at the front and rear door posts and ahead of the rear fender openings. From the spring mounts out to the bumpers you really won't gain much, as the spring perches are where the twist load is fed into the structure.I'm following and thinking since I do want to potentially get the body off of the frame that I'd weld the rocker tubes/channels to the frame and put extra mount plates of some kind welded to the tubes/channels and then bolt the body to it using some form of high durometer mount (hockey pucks come to mind). When you say "at the front and rear door posts and ahead of the rear fender openings" are you talking plumbob like underneith the posts or mearly in the basic area?
67808
This is the best picture I've seen of the front of the frame, you can see the double front body mounts that are up under the firewall.
67809
This is the "Hemi Under Class" Arias landspeed car which uses a heavily modified stock frame. You can see the stock front body mounts and the rear suspension (it's a 3 link) lower arm attach points. If I did a straight bar like they did, it's cut straight through the floor pans since they are basically channeled over the frame and the bottoms of the pans and the rockers are nearly level with the underside of the frame! I'd curve out from the front body mounts to running either just inside the rockers if I kept the double rocker channel on the body (I'm guessing Ray is nodding here) or physically replace the inner rocker (it's around 4-5in inboard of the outer rocker) then curve back to the rear mounts like they did. Likewise I'd use the space under the seats to tie the outer tubes/channels to the X and probably triangulate them as much as possible without cutting the floor pans up to make it fit.


The cross bar between the A pillars will help if the firewall is a good distance ahead of that area. I'd make it out of the largest diameter tube you can possibly fit behind the dash. It can be very thin wall to save weight. Think about welding a driveshaft across there and you'll get the idea... One way to improve the stiffness of the floor is to fab the seat risers as boxes (out of thin gauge sheet) and run them from rocker to tunnel. If you're keeping the rear seat, build a sheet metal riser that goes WAY up and run a thin foam seat bottom. Weld in a thin plate to seal off the rear seat back bulkhead. None of this has to be very heavy, and dimpled holes can reduce the weight further at little cost to the structure.In general terms, how far is a "good distance" in front of the door posts? Mines probably 18in or so since depending on if you include the cowl plenum for the fresh air vents and even so it's shaped like a 'K' with the bottom left leg cut off since the toeboards come up then it's vertical to the bottom of the hood if you count the leading edge of the plenum but the backside of the plenum angles back towards the base of the windshield. The top is bolt on and it's deeper near the sides of the car and shallower in the center in terms of vertical depth.
I already have a stamped channel that runs from inner panel to inner panel between the door posts but it's probably extremely noodle like. Remembering an article I read on the new for '97 Park Avenue, they put a Mg strut all the way across the dash area and tied the dash mountings to it too reduce squeaks and rattles, I'm guessing I could the same here. The only major hickup with that idea is the steering collumn as the existing span piece runs just above the centerline of the steering collumn. Can I put a boxed notch in the tube to route the collumn through since the only way to get the instrament cluster out is to drop the collumn as far as I can tell?
67810
The seat box idea is a great one for the front since the floor actually kicks up for the seat mount area leaving space underneith if I don't eat that up with frame bracing. Between the two, cross bracing from the outer tubes/channels to the X and a seat box, which is more benificial? I'm also going to have to massage the area since I'm putting late model Eldorado seats in and they have different mounts. The rear seat area similarly kicks up then you have the rear axle kickup at the back with these braces to the parcel shelf. Do I need to further panel it in?
67816

Thanks,

[exwestracer]

If you have to put a lot of bend in the "rocker bars" for the frame, or you can't make them very tall, I'd forget them all together and concentrate on tying the body soundly to the frame. In that case I would concentrate the structure in the floor sheetmetal, not the frame (too much floor cutting to make it work). The door post mounts should be as close as "practical". You don't have to plumb them exactly... Since there is NO frame structure at the rear door posts, I'd look at running some "outriggers" off the frame that you can tie into the floor where the rear seat step up is.

That parcel shelf area is already closed off pretty well. See if there is any practical way to tie the top spring (coilover?) mounts to the inner fenders and brace them up into the roof pillars.

[CTX-SLPR]

Looking at the floor pans lastnight made me reevaluate my plan to have both a boxed rocker and the rocker bars. There is only enough realestate for 1.5-2in tubing in that area unless the inner rockers come out. Based on what you were saying Ray, it's better to stiffen the body than to stiffen the frame all others being equal. That being said, there is room to use 2in tubing to tie the fronts of the rear fender opening and the rear door post (both of which have a cross body underfloor brace) to the the frame with more body mounts. I can only really find the front doubles (4), spring cross member (6), and a pair on each side of the trunk pan (10) body mounts on the whole chassis. Seems to me like there is something to be gained by atleast tieing the rear door post to the frame if not it and the front of the rear fender opening. The front door post would be hard for ground clearance reasons as the bottom of the cross body under floor brace is basically even with the underside of the rockers already.

Rear springs are on the crossmember you can see in the Arias picture and work on the lower control arms themselfs. For now they are going to still be there as is just not enough real estate to put a different kind of rear suspension back there other than maybe a Watt's linkage instead of a Panhard rod as a lateral locator. Putting coil overs where the shocks are in the back would chew into wheel and exhaust room pretty good as they are outboard of the frame rails. I've thought about putting them in the middle just inboard of the frame rails in the axle kickups but that might be too narrow in such a heavy car. There are body mounts on that crossmember so I could see if tubing would fit throught the sail panels up to the roof or atleast the metal structure inside the sail panels.

[exwestracer]

Are you mini-tubbing the rear or cutting into the wheel tubs at all?

If not, the inside of the panel of the inner fender should line up pretty close with the outside of the frame rail. I'd plate that vertical face of the wheel tub down onto the frame rail. Again, it doesn't have to be heavy gauge, and it should be as long as possible. That will transfer the spring load up the inside of the tub and into the C pillar for almost zero added weight. There should be some good structure right there in the trunk hinge supports.