New car has no cage, I know the 914 tends to flex and that the factory welded steel roofs on the 916 for what I guess was to help stiffen them, or was it for other reasons??

The 'glass top should help a little, but I was wondering about attaching tubes or the like from the factory roll bar to the windshield hoop to stiffen the car up....

Has anyone played with this before??

An X brace would help alot to stiffen things up, but how stiff is the windshield frame anyway?

Geoff

Mike,
Don't have historical data on the 916 but I plan to use engman's lower bracing if I don't use a cage. Here's a hack solution!
Felix

I saw this on one car. Don't know how functional it would be.

LOL same pic

And this is on an original 914-6 currently at ~600 on eBay!
Felix

Mike,
It's so funny to ready your threads like this becuase you usually post about something that I'm actually scratching my own chin about. I've thought about doing a cross-brace like this myself. I was thinking about maybe doing a box-steel 'X' of light-guage steel and glassing it into the undersid of the roof. Then tie it in to the targa bar and windshield frame w/ removable plates.

Now that I've got a new daily driver I'm just going to do a cage though.

-Ben

this is the brumos gt

front

back

I thought I read somewhere that the 914, new and un-rusty, was designed to be just as stiff as a 911 coupe. Has anyboldy else heard of this ?

Michel Richard

I know even with an 8-point welded cage, the tub still has "some" flex, because I have a skin targa top attached with dzus fasteners and it squeeks and rattles quite bit. That tells me there's a little motion going on. I figure it's the A-pillar (windshield frame) since the cage is tied into the targa bar. The A-pillar on any roofless car or even a sunroof car is inherently gonna flex.

Maybe 911 Targa.......

On the 914-6Gt pages the Gt in the Collier collection shows a picture of a reinforcing plate added at the base of the windshield frame. along with the bolted X-braced Gt top.
The factory must thought it would help ?
new car and the wheels are turnin already

Like Geoff already mentioned, the windshield frame is pretty weak
The first factory race cars cracked their windshields because of the reinforced roofs
Later they also reinforced the pilars of the windshield frame

So... get Engman's kit and be done with it...

Cool. That's almost exactly what I was talking about. I expected the cross-brace to be lighter and the tie-ins to be beefier, but that's the general idea I was talking about.

-Ben M.

I've "heard" they're much stiffer.

Troy put the A-Pillar mod on Gint's car. Came out awesome. These things crack a-pillars, especially when autocrossed or raced.

When I restored my 911 the a-pillars were cracked. The factory did the reinforcement to prevent cracking. A-pillars are weak, weak, weak.

Here's a couple of pictures that Andy came over and took... (actually, the camera was pointed directly at the light so it didn't like to focus properly.)

nudder

Never push the car from the a pillars, I have my two boys, my wife and most of my friends "trained" by now!

I usually drill some holes in A-pillars and shoot in some expanding hard foam insulation. Not the home depot crap. But, the commerical stuff. Underneath the weather strip area and behind the A-post trim gets it in all the voids. It seems to add some strenght to the A-pillars and cowl area. Some cheaters back in the old days use to shoot foam into the sills, A-post and frame rails on some cars to strenghten the car a little. Not me, but other cheaters. Undercoating usually did a good job at hiding it under the car. At least that's what I heard.

If you add steel into the top you are hanging some major wieght up high in the chassis. It will have an affect on how the car feels. I think you would be better off laminating the stock top with some sort or honeycomb material and fiberglass. Anything you could do to make it stiffer with out using the steel the better.

The two part expanding foam works well. I know Ford is using it in some of its new vehicles. Make sure the chassis has no rust in the frame rails. Then fill with a commercial two part expanding foam. This will do wonders for the frequency of the chassis.

Engmans kit is a great addition for the average joe. Most cages that are put in cars only add weight and safety. Hopefully when I get some time after the first of the year I am going to be doing some torsional testing on my chassis. I just have to finish the chassis jig first.

You guys surely jest....

The A-pillar weak. Please.

Have you ever cut a windshield off? There is a bunch of steel in there. Several thicknesses of sheet metal. Potentially, in a hard collision, the windshield frame may try to crack at the point where the rigid frame meets the rest of the unit body.

Those "stiffening plates" In my opinion were put there to reinstall a previously cut-off windshield frame.

Our cage has been tied into our windshield frame for over 16 years of track time and our windshield frame has not cracked off.

Now, I admit that there are some high stress concentrations in the windshield frame/cowl/fender area. We noticed that the fender/cowl interface is under considerable stress when we welded the cowl/fender joint and it kept cracking/moving.
We ended up cutting that area apart and replacing the steel in the area, but the area continues to do strange things especially at the sharp corner at the rear edge of the front trunk lid.

Yes I have cut the windshield off. The windshield and cowl wiegh in at less than 10lbs. There really isn't a lot of material in that area. There are several pieces of material that come together around the door jambs.

Notice the redline is a broken spot weld. You can see through to the fuel tank area. That panel is 1 peice of .050 sheetmetal. No top just an open unreinforced box. There is very little structural intergrity to this area. That is why these chassis twist so bad. The lower A pillar is really strong, not necessarily stiff. It does form a boxed section but the box is vertical and not really tied into the rest of the chassis. The reason it is so strong is because the doors have to hang off them.

The Blue line shows the multiple peices of metal that come togther in the area of the A-pillar, but they just connect like post. Vertical with the rest of the chassis hanging off the post. Very little reinforcement on three sides.

Agreed.

So how does the previously referred to "reinforcement plate" welded into the lower A-pillar address any of this or help with the area where you show a broken spotweld?

It doesn't. What that patch does is, connect the windshield frame to the chassis a little better. It helps keep the windshield frame and the cowl/firewall flexing at the same rate/amount.

looking at the pictures of the cracked windshield frames, it seems like a more solid roof might increase the cracking at that point......

I think for a street car with no cage, if there was a way that one could make their doors more of a "stressed" member, that would really help stiffen the car. From what I understand the door jams and the hinges are pretty weak as is, so things would have to be reinforced.

It would be interesting to measure how much the door gaps increase/decrease while driving.

Nope. And don't call us "Surely"

All I can tell you is "every single" restoration project I've tackled (2 911's and 2 914's) has had cracked a-pillars. We're currently looking at a 914-6 on a rotisserie (sic) that the windshield posts move more than 1/4".

Factory GT's have the stiffening plates welded in and these are cars that have a clear history of "never" having the windshield cut off. I'll cite the aerospace engineer that was looking at the car when we were contemplating the factory stiffening kit... "It's funny how many good-ole boys think they're smarter than the factory engineers".

So, I'll go by experience with all my 911's and 914's. A-Pillars crack. A picture is worth...?

need more visual proof?

The problem with the 914 chassis is there are way to many disconnected and incomplete box sections. Look at the front trunk area and fuel tank area. Uncapped boxes, disconnected structures. Look at some of the newer chassis and you can see how things are much different. Look at the Carrera GT. I realize that is a carbon monoque but look at how the box is built. Look at some of the Audi Unibody chassis. Very stiff. Look at the new Ford GT.

Many people are working on a way to configure the doors to be part of the intergral chassis structure. Unfortnately, keeping the unit light wieght and durable has become a very large problem. In theory you could design some sort of door with multiple pins and moving spacer blocks that would put pressure on the chassis at the same points that the cross bolts go into, but that would be a major engineering and manufacturing hurdle. Just imagine what happens when the system fails and grandma is trapped in here Cadillac. Car manufactures would go out of business from lawsuits.

Check out the new issue of Grassroots. They have a fairly lame article on Chassis stiffness based on their 914 project car.

Good Ol boys can know more than Engineers. Just because you are an engineer doesn't make you God's gift to design and production. I know engineers that couldn't find their way out of a brown paper sack, but they could tell you torsional stiffness of that paper sack, in theory. Obviously there are exceptions to the rule. It all depends on the situation. I agree most don't know more though.

Hmmmmmmm... me understands but (there's that "but" word), I would think that a Porsche Engineer assigned to the factory race team might be the "exception" to your paper sack rule. They engineered them to 1st place in GT class at LeMans after all. I'll be trusting them until I get some cold hard facts to the contrary (a "fix for a windshield being cut off" would not be that. Sorry John ). Especially when I've had first hand experience with so many cracked windshield posts. It sounds like we're on the same page with this.

We did some pretty solid research on the Frame Rails and the Factory Stiffening Kits and I feel there's enough evidence there to contradict a lot of the "here say" going on here. We had a Boeing engineer come look at a bare -6 tub and describe what he would do to stiffen the chassis. Without even seeing one, he literally drew a factory kit on the tub. He looked at the frame rails (longitudinals), rear shock towers and the tranny mounts and came to his conclusion. He also mentioned stiffening needed for the rear suspension console. It was pretty eerie watching him lay it out.

All in all, there's a lot that 30 plus years can teach us about the strength and longevity of our tubs. They were an engineering marvel for there day. I'm sure there's some things we can understand and see now that they didn't have the time or wherewithal to address back then. A lot of times they simply engineered them to last 24 hours. The 914 was a very short-lived project for the racing team.

On John's behalf, the fact that his cage was tied to the frame and probably the rest of the car through the longitudinals would very much answer the question as to why his windshield posts never cracked in 16 years. I would doubt there would be any flex in the frame with it tied in so well.

So... "Good-ole-Boys", I'm all ears (really I am). If there's some better way, let it be heard.

I'm not an engineer, but I get a laugh out of some of the engineer comments from time to time. You can think what you want but here's my take. Not all engineers are good at designing structures, but ALL of the BEST structural designers are engineers.

-Ben M.

Like Eric mentioned... that is exactly why your windshield DIDN'T crack...

We all agree that our targa cars flex at the top.
That leaves 3 pieces: the targa bar, the roof and the windshield frame
The targa bar is a pretty surdy piece (it's proven to function as a roll bar)
The windshield frame is quite flimsy... you can move it just by hand and you don't have to put much force into it.
Put a solid connection between those two (a reinforced roof) and which one do you think will end up suffering from the stress?

Just wondering how much reinforcing the rails would help. The problem is lack of torsional stiffness because the car has two rails on the outside with a floorpan, and not because the rails are too limber. Due to the design, stiffening the rails will help a little but not that much.

If we assume the height of the rails can't be changed (look at an Elise for example), the only other solutions I see is to either mount a steel roof or create a backbone (like the C5/6 Corvettes, Lotus Elan). I don't think there is enough space between the seats to fit a backbone. As an alternative to a steel roof, putting crossbraces between the targa and the windshield frame is prolly not such a good idea in the long term cause you create serious stress concentrations. Installing a full roof so the loads are more evenly divided between the A pillars should be better.

I've been thinking about how to increase torsional stiffness for a while now and I think I'll just live with what I have ... I ain't putting no stinkin' fixed roof on my 914.

Rather than attaching your top, consider some tubing.

I put this stuff in my car to help it handle a big engine, it's not for safety. The increase in stiffness is nothing short of amazing. The tubing continues, uninterrupted to the front and rear.

Let me know if you'd like to see more.

If that goes into the front trunk and the engine compartment, I wouldn't mind seeing pics of those areas.
Thanks

This is the best one I have right now (I'm on the road).

If you look carefully at the bottom of the picture, just behind the trunk spring mount, you can see the welded mount for the 1" tubing coming from under the dash. If you look across you can see the opposite tube going to its mount. If you look in the front, you'll see the two 1" tubes residing in that space.

All the tubing in this car is straight in order to keep rigidity to an absolute maximum.

I can take additional pictures when I get home later this week. I have good pictures of the engine bay and rear trunk.

engine bay:

rear:

No Eric. That will do. A picture is worth a thousand words as they say.

I tied the windhield frame into the cage at the top of the windshield frame probably 16 years ago.

The windshield frame moved enough on the track that it would actually rub on the cage and would otherwise create noises. During that season, I decided to tie the frame into the cage to prevent the squeaks/rattles. After doing that, I pretty much forgot about it.

The areas that we experience cracked paint is normally in the same area, but further forward. We also (in our infinate wisdom) welded the joint between the fender and cowl. This area would routinely crack the paint until we cut out the joint and replaced it with sheet metal. Our cowl and fender now form a 90 degree sharp corner and this seems to be a stress concentration as that corner is where our paint will crack now.
(I am convinced that this is the reason for the plastic welting in the joint to begin with). In other words, I believe Porsche knew the fenders/cowl/windshield frame moved too much to weld all that together and created the "expansion joint" to address the issue.

I have noticed that the cracks in our paint starting at that sharp corner (stress concentration) start only after a severe jar (dropping a wheel off the edge of the track into a HOLE or hitting VERY VERY bumpy alligator strips). These severe "jars" are HARD. Hard enough to bend a strut (which I have actually done on one occasion).

you could always follow what they did on some of the GTs ...

here's the lufthansa GT ...

Andy... awesome picture. Where the hell did you get that? (we know you didn't take it because, a. it's in focus, and b. you'd probably be about 1 year old when that was taken)

With this thread I was wondering about how they made the bracing to fit so precisely. Looks like they welded it up first and then glassed it to the top. Perfect timing (for me anyway). Thanks for the pic. Any more?

John, You bring up another good point. There's a lot of people on the board that weld that cowl joint up. I "think" Brad's #0004 had them welded shut as well. From your real world experience it seems like those really "need" to be there. Not a good idea to weld them shut.

Great pic . , link? Any more , Havent seen that one before.

What about glassing in some square or rectangular stock to the targa top (or just a stiffening U/square channel and or X brace with glass and matt alone), one on each side by the window seal, fore to aft. to give some spine to the top. The front latch is pretty stout to pull the top to the windshield frame (compared to the rear). On the rear change out the hook and pin latch mechanism for a straight down pin type latch (like the rear trunk), then you could set the pressure to preload the rear seal, or even have the pins be adjustable fore aft within a channel to preload push or pull needed to keep things square. That would hold it a bit more steady than what is in there now and could be made to be fairly light. Obviously not a super strong approach, but could use what's there with a couple tweaks for a super duty street car.

my windshield frame will wiggle easily when I grab it and pull it back and forth, I too was wanting to reinforce the A pillar area to help. If I do this, what should I use for material, some flat bar stock bent and welded, some heavy gauge sheet stock with some beads?

What about where the cowl meets the windshield base, can this use some extra support, then tie to the upright supports as I talk about above (like using 1", 1/8 flat bar stock), from windshield support to windshield support, effectively making a "U". You could bend the bar stock to get close to the windsheild base and follow the contour, wrap around the corner to the upright supports. This could be done in the front trunk compartment, or under the dash in the cabin - so it would be hidden.

Guys you are missing the main problem. Not enough complete box sections in the chassis. If you can complete more boxes, increase the cross section of existing boxes, you will gain eponentially more stiffness than you would if started making all of these band-aids.

Look at the early aluminum Monocoques. They used very few tubes and many panels. You could take the factory tunnel out and put in one of a much larger cross section. You could increase the size of the factoy rocker boxes. Which are significantly stiffer than just about anything your are going to add or replace it with. Make a huge backbone through the car. This is probably the best thing you could do for the strength of your chassis.

You could build a very light weight frame of steel tubing for a seat that basically fills up the interior of the car. Make places to two seating areas and skin the whole thing in aluminum. Attach your frame to the basic 914 chassis and bond and rivet the aluminum sheeting to the frame and the chassis. This would increase your torsional stiffness incredibly. Yes it will be a massive amount of work. Would it be worth it. Absolutely.

from what I have read here, it now seems to be a bad idea unless you reinforce where the windshield frame meets the cowl area (A-pillar)....

the factory roll bar is stronger than the windshield frame, if (more like when )the car flexes, the forces are going to find the weakest link and that would be the windshield frame...with enough force, the windshield could crack....I'd like to know if the factory 916s had this problem, but I highly doubt many of them have 100k+ miles on the chassis and still being daily driven

Sorry I wasn't clear, I was meaning doing the windshield frame support and top reinforecement together, again does that sound reasonable.

Brett,

Where would you finish completing boxing of the 914 chassis, I want to visualize what or where to do this as you are stating

Travis, with the exception of your pin on the rear section (they bolted theirs in) it sounds like your talking about a factory GT setup. They also glassed theirs to the roof.

Gints a-pillar mod was sheet stock (10 gauge?) and followed the GT pattern. Check the photos on Jon Lowes site for more on that. It will be completely hidden once the interior trim is added. He commonly grabs his frame to get in and out of the car.

I think everyone's on (to) something here...

Brett seems to be talking about an all out FP type car or more.
John's car has a welded in cage that probably negates some of these issues.
Gint's car is a -6 that will have a mild-hot motor, will see some track time and won't have a cage.
Mike - ? I don't know what your plans are for that car.

We've discussed the issues of bracing the a-pillars and the roof ala a GT but really haven't addressed the question. "How much stiffer with solid or tied in roof?"

I "think" solid would be noticably stiffer (I only have a 911 to reference and old stories about the 916). Tied in like a GT would probably be noticably stiffer as well. But, as Brett and/or others have mentioned, most of your rigidity will probably come from strengthening the longitudinals. Might as well keep the targa top as it is and get an Engman kit. If you did weld on a solid roof you can still have problems if you don't address the flex that John points out (which probably means full tube, seeing as how he has a full cage and still has the flex). I raced my 911 without a roll bar or a roll cage for quite a few years and when it came time for the RS conversion the a-pillars were cracked. It's a non-sunroof coupe.

If someone were to come up with a 916 roof blank I think I'd consider it

Ohhh! A 962. Me likey.

(It is a 962, right?)

-Ben M.

Just got my engman kit last week, and a welder yesterday. Just trying to figure out what things I can ruin errr um strengthen heh.

Thanks for the metal size for the uprights, I will definately do that just for the fact it's so flexy on this car.

Ok Eric,



Back to the topic....

In my opinion.... simply tying the roof into the existing windshield frame and the Roll Bar (without the benefit of a full cage) would only marginally stiffen the chassis.

just my $0.02

Eric you could use it on any type of car but it would do away with the regular up right seating and lay the driver back in a very Formula car"esque" seating position. Might be a problem for tall drivers. Look at the seating styles for some of the GTP cars, 962, McLarens, etc.

Travis
Starting at the front of the car, front trunk, big open box. (not really a big concern because it is just holding up the sheetmetal). Fuel Cell area, under windshield cowl. Open box with three panels. all connect to different pieces. None share a common joint. Both A-pillars just standing out in space. Partially bridged by the windshield cowl, sorta not really. The manufacture basically wanted the joint between the cowl and firewall to form the majority of the strength in that area. To bad it didn't work. So basically you have the strut towers connected to these beefy pieces of metal (the strut towers) with nothing on either side of any substance. So they pivot back and forth. The A-pillars flex in and out.

The tunnel is connected to one little piece of .050 sheetmetal with a couple of spot welds. That tunnel dead ends into a flat piece of material with nothing behind it but open space. The firewall wiill flex in bending and torsion. The framerails them selfs aren't all that bad. they are really stiff in bending and would be in torsion if they were connected together.

In the back the tunnel is connected another piece of .050 sheetmetal. This piece is backed with another peice of similar thickness but it is only connected in a couple of places. The firewall is folded over and spotwelded to the front engine shelf. That provides some stiffness but not a whole lot. Engmans kit helps here a little but all it does is add thickness to the one piece of sheetmetal. What it should do is build a box that comes off the firewall and is connected to the tunnel with another boxed piece. This would cause seating issues for most people.

The rear frame rails are braced by on pre-bent peice of .050 sheetmetal (the inner fender panel). So what happens when these are loaded in torsion, as the factory suspension points loads the rear suspension, The box tends to twist in and up. Basically trying to roll up towards the engine. Now a shaped panel that fits in the corners of the engine bay would do wonders for increasing the box section of that area. Adding the 2inch .043 wall tube underneath the engine lid mounts and 3in x.043 under the rear engine shelf would make noticable difference in a stock vehicle. As would bracing from the fire wall to the rear suspension mounts.

Remember most manufactures have packing issues and cost issues that come into play in deciding where stuff should be. So these issues might have gone away had the model been updated several times.