Fastback inner rockers

[Mach1 Driver 560]

7 hours ago, fvike said:

Did you see all the videos? There's a playlist with 13 episodes. Ep 9 shows retesting at Multimatic for damper tuning. The Videos show re testing on the 4-post rig, and the Multimatic people at the track for real world testing. The job was done right. No, they do not give away their test results, just as they don't with the spring rates and the damper valving. That way a competitor would get them for free. I'm sure the hours at Multimatic wasn't free. 

I'm telling you, all you have to do is drive a car before and after getting SFCs installed. If you have, we wouldn't have this discussion. It is a night and day experience. 

The Barracuda and the Mustang have differences, but they both are unibodies with separate front and rear framerails. How supposedly one benefits form SFCs and the other don't, you'd have to explain to me. Well, apart from removing the shock towers.

All the testing done in the "garage test" is meaningless according to post #24 on that tread.

I watched the three videos you had linked. If there were more I didn't see them and based my conclusions on the three. I'm glad to hear they went back for more testing. But since all the changes were on the car they couldn't tell which changes did what. I have stated several times that members have added SFCs and feel real differences. All I'm stating is that in Torsional rigidity test: 67 coupe, Stangnet.com  in test 1, 2 and 4, that SFCs did zero. He made some significant additions to the SFCs. This is From post # 60 on page 3, just prior to test 4:

Jacking rails
In preparation for the next round of testing I have finished the jacking rail/sfc assembly. The SFCs are from Heidts and the jacking rail kit came from Global West. I purchased it from Summit Racing for about $130. Since it was made for Global West SFCs there was quite a bit of modding and fabrication required. The 1 inch square tubing welds to the rocker panels. The portion that extends forward will tie into the front torque box. The kit appears to be designed for installation on a car without front torque boxes so more mods were necessary.

 

You will see in the summary that this still did zero and that surprised the heck out of me.

As far as the meaningless comment: 

from post #24 page 2

CraigMBA is relating what his wife who is a structural engineer said: You can measure deflection at the core support, but the measurement is meaningless because you don't apply any torque there anyway.
I think something was lost in the translation from wife/engineer to CraigMBA. It isn’t meaningless at all. Who cares if torque isn’t applied there? You have to measure somewhere and this is a before and after comparison to give a percentage, not an absolute defining measurement. Look at the setup. The car has jack stands at three points: RR, LR, and LF. The RF is free to flex so the logical place to take the measurement is at the forward most RF corner where you will see the most deflection.

This car has significant changes but it is still a 67 Mustang. It has a newly constructed firewall for a big engine, the shock towers are removed, there are braces from the cowl to the front frames, and he has fabbed "export braces". The point is that he is measuring deflection before and after he makes changes to see how much (percentage wise) the change affects how much the chassis flexes. 

Here is the summary from post 79 on page 4:

Torsional rigidity test: 67 coupe, Stangnet.com
The improvemens in % are the relative changes compared to the previous step, so the for the Export Brace: (0.64-0.48)/0.64 = 25%
Test1 (post 1 page 1)
Baseline 0.66"
Factory crossmember 0.64" 3%
Export Brace 0.48" 25%
Monte Carlo bar 0.48" 0%
Passenger torque box 0.48" 0%
Subframe connectors 0.48" 0%
Subframe X brace 0.475" 1%
Test2 (post 1 page 1)
Baseline 0.67"
Cut shock towers 0.69" -3%
MII cross member 0.63" 9%
Subframe connectors 0.63" 0%
Subframe X brace 0.61" 3%
Test3 (post 23 page 2)
Baseline 0.62"
rear mounting point forward 0.52" 16%
front mounting point rearward 0.35" 33%
Test4 (post 66 page 4)
Baseline 0.44"
Firewall to subframe braces 0.42" 5%
Weld braces to aprons and
shock tower panels 0.41" 2%
Diagonals ("Export Brace") 0.36" 12%
SFC 0.36" 0%
SFC X brace 0.36" 0%
Jacking rails 0.35" 3%
Rear seat divider 0.31" 11%
Test5 (post 75 page 4)
Baseline 0.35"
rear mounting point forward 0.19" 46%
Removed SFC X brace 0.19" 0%
Removed seat divider 0.23" -21%
Removed seat pans 0.23" 0%
Front mounting point forward 0.40" -0.74%

Most effective changes
test
1       Export Brace 25%
4       Diagonals 12%
4       Rear seat divider 11%
2       MII cross member 9%
4       Firewall to subframe braces 5%
2       Subframe X brace 3%
1       Factory crossmember 3%
4       Jacking rails 3%
4       Weld braces to aprons and shock tower panels 2%
1       Monte Carlo bar 0%
1       Passenger torque box 0%
1,2,4 Subframe connectors 0%

It's interesting to see that the simplest mods seem to have the greatest effect. But, as mentioned before, one should keep in mind that although some mods showed little to no improvement in this particular test, it doesn't necessarily mean that they are therefore worthless.

 

 

[Mach1 Driver 560]

7 hours ago, RogerC said:

Having dabbled in FEA with a few different CAD programs it would be easy to see the difference.

Yes those programs have been around for years and I have used them also. However in my opinion it would not be "easy". In the CAD programs I have used, first you need to make an accurate 3D model of all the chassis parts. Then you have to connect them together in an assembly (with lots spot welds), and then you can do a structural analysis. I wonder how many thousands of hours that would take? Ford never did it because the technology wasn't around then. Maybe Dynacorn could do it now, but they are in the business of supplying replacement parts, not redesigning the car. But boy it sure would be nice.

[RogerC 136]

The modeling wouldn’t  be so easy but the FEA wouldn’t be that difficult either and the results would easily show the differences. Was meaning to post later but my oldest daughter went into labor with our 2nd grandchild. Bored and posting from the waiting room. 

[fvike 173]

19 hours ago, Mach1 Driver said:

I watched the three videos you had linked. If there were more I didn't see them and based my conclusions on the three. I'm glad to hear they went back for more testing. But since all the changes were on the car they couldn't tell which changes did what. I have stated several times that members have added SFCs and feel real differences. All I'm stating is that in Torsional rigidity test: 67 coupe, Stangnet.com  in test 1, 2 and 4, that SFCs did zero. He made some significant additions to the SFCs. This is From post # 60 on page 3, just prior to test 4:

Jacking rails
In preparation for the next round of testing I have finished the jacking rail/sfc assembly. The SFCs are from Heidts and the jacking rail kit came from Global West. I purchased it from Summit Racing for about $130. Since it was made for Global West SFCs there was quite a bit of modding and fabrication required. The 1 inch square tubing welds to the rocker panels. The portion that extends forward will tie into the front torque box. The kit appears to be designed for installation on a car without front torque boxes so more mods were necessary.

 

You will see in the summary that this still did zero and that surprised the heck out of me.

As far as the meaningless comment: 

from post #24 page 2

CraigMBA is relating what his wife who is a structural engineer said: You can measure deflection at the core support, but the measurement is meaningless because you don't apply any torque there anyway.
I think something was lost in the translation from wife/engineer to CraigMBA. It isn’t meaningless at all. Who cares if torque isn’t applied there? You have to measure somewhere and this is a before and after comparison to give a percentage, not an absolute defining measurement. Look at the setup. The car has jack stands at three points: RR, LR, and LF. The RF is free to flex so the logical place to take the measurement is at the forward most RF corner where you will see the most deflection.

This car has significant changes but it is still a 67 Mustang. It has a newly constructed firewall for a big engine, the shock towers are removed, there are braces from the cowl to the front frames, and he has fabbed "export braces". The point is that he is measuring deflection before and after he makes changes to see how much (percentage wise) the change affects how much the chassis flexes. 

Here is the summary from post 79 on page 4:

Torsional rigidity test: 67 coupe, Stangnet.com
*Summary*

It's interesting to see that the simplest mods seem to have the greatest effect. But, as mentioned before, one should keep in mind that although some mods showed little to no improvement in this particular test, it doesn't necessarily mean that they are therefore worthless.

 

 

I think the meaning of post #24 is that the method applied is wrong. As in adding SFCs will do nothing for radiator support flexibility. I can very well believe that. But it will do a lot for the chassis. And the method applied misses that. The setup is the problem. The frame is supported by 3 jack stands. The whole of the chassis isn't available for flexing because it is locked in place. He is just measuring flex in the front frame rail, not the whole unibody. When the shock towers and Monte Carlo bar was in place,  that load was spread across a larger area. When removed, and the same force applied will show more flex because the area it is applied to is less (limited).

The 25% torsional stiffness credited to the MC bar isn't because the MC bar is so immensely strong. It's just a piece of steel. It is because the MC Bar allows the frame rails to be connected to a larger area.

He should have put the car on it's suspension, but with strut rods instead of shock absorbers and springs. That way, the frame itself would not be supported at any given point, and free to flex all the way from front to back.

[Casgar 21]

Late to the game, but I find this thread really interesting.

 

After reading thought it all my take from it is that SBC and inner rockers by themself only help reduce sagging, and by connecting the two sides with each other via a one-piece seat riser or X-brace you reduce the twisting.

This might be a dumb question, but wouldn't a piece of say 1/4"x3" flat steel or L-profile welded to the inside of the rocker insead of an actual inner rocker add just as much sagging reduction as SFC or inner rockers, but be much less intrusive inside the car and underneath it?

As for the twisting motion, replacing the rear leaf spring supension with something else might actualy help too. I saw this video yesterday and it only reinforced my plans to building a torque arm suspension. What's your thoughts on it?

 

[Mach1 Driver 560]

That's a good video. At the beginning at 0:28 he talks about beaming and torsioning testing and that most SFCs are just good for "ballast in the boot after you cut them up". If aslanefe, our resident aeronautical engineer will chime in here (I'm an electrical engineer and didn't deal much with structures) he could tell you how thick a single plate of steel would need to be to equal the strength of a rectangular steel tube. I'd rather not revisit my college days in my retirement to look that up. If you notice, starting at 3:35 they welded in a hat shaped beam, turning it into a rectangular steel tube. Good stuff, and it is further verification of the torsion test shown much earlier in this string. SFCs without bracing from side to side do next to nothing for torsion.

[RPM 1,190]

4 hours ago, Mach1 Driver said:

That's a good video. At the beginning at 0:28 he talks about beaming and torsioning testing and that most SFCs are just good for "ballast in the boot after you cut them up". If aslanefe, our resident aeronautical engineer will chime in here (I'm an electrical engineer and didn't deal much with structures) he could tell you how thick a single plate of steel would need to be to equal the strength of a rectangular steel tube. I'd rather not revisit my college days in my retirement to look that up. If you notice, starting at 3:35 they welded in a hat shaped beam, turning it into a rectangular steel tube. Good stuff, and it is further verification of the torsion test shown much earlier in this string. SFCs without bracing from side to side do next to nothing for torsion.

I too liked the video (although coming from the guys who love to put Mac P strut front ends in early Mustangs???). The weight of a flat bar to equal a tube for a sfc will greatly outweigh any space savings.

[Casgar 21]

53 minutes ago, RPM said:

I too liked the video (although coming from the guys who love to put Mac P strut front ends in early Mustangs???). The weight of a flat bar to equal a tube for a sfc will greatly outweigh any space savings.

Is it not only the height and thickness of the material that is relevant when calculating beam strenght? As such, wouldn't a piece of flat steel with the same height as an square tube and the thickness of both vertical sides of the same tube have the same beam strenght? As an example, my local hardware store have 50x50x4mm tubing weighting in at 5,64kg/m. A piece of 50x8mm flat steel weight 3,14kg/m (pardon the metrics).

Also, replacing SFC with flat steel isn't really practial, I agree, since it would't have any side support and would most likey bend sideways. But I don' really see why you can't just replace the inner rocker with a piece of flat steel instead, unless it have some other function beside beam stiffening. The flat steel would leave me with more room between the rocker panel and transmission tunnel to put big comfy seats in between compared to inner rockers. Then again my reasoning is based on me closing my eyes, visualizing a car model in front of me and then bending and twisting it with my hands. :)

[Vicfreg 771]

The video is cool, but he is slightly off calling any '60s or '70s Ford Chassis a Monocoque chassis. It is a actually more of a semi-Monocoque chassis, as it has sub-frames, rockers, and torque boxes.

A true Monocoque would be a modern Formula 1 race car with an integrated chassis and body carbon fiber skin, where the overall skin of the structure creates the actual structure of the car.

There are a number of SAE standards that measure the modulus of stiffness (torsional rigidity) for automobile structures and materials.    I am an old Mechanical Engineer, with a bad memory, so may be wrong, but in my simple mind, hooking the front and rear uni-body assemblies together has to provide some level of benefit.    I can feel it in the cars I have driven, mostly in the way the car handles.  

Anyway, keep welding stuff in..it's fun at the very least.

 

 

[aslanefe 333]

7 hours ago, Casgar said:

Is it not only the height and thickness of the material that is relevant when calculating beam strenght? As such, wouldn't a piece of flat steel with the same height as an square tube and the thickness of both vertical sides of the same tube have the same beam strenght? As an example, my local hardware store have 50x50x4mm tubing weighting in at 5,64kg/m. A piece of 50x8mm flat steel weight 3,14kg/m (pardon the metrics).

Also, replacing SFC with flat steel isn't really practial, I agree, since it would't have any side support and would most likey bend sideways. But I don' really see why you can't just replace the inner rocker with a piece of flat steel instead, unless it have some other function beside beam stiffening. The flat steel would leave me with more room between the rocker panel and transmission tunnel to put big comfy seats in between compared to inner rockers. Then again my reasoning is based on me closing my eyes, visualizing a car model in front of me and then bending and twisting it with my hands. :)

Deflection is inversely proportional to the moment of inertia of the shape (profile/cross-section of your material); not only thickness and height.

According to some online moment of inertia calculators, your example tubing has moment of inertia of 261525 mm^4; your example flat bar has moment of inertia of 83333 mm^4.

So your example flat bar has less moment of inertia (about 4 times less) and it will deflect more under same load.

You need about 50x25mm flat bar to have almost same moment of inertia of your example tube.

On your previous post you said "As for the twisting motion, replacing the rear leaf spring suspension with something else might actually help too." Twist happens between front and rear suspension attach points, so you have to tie right front to left rear, left front to right rear to reduce twist; replacing rear suspension to 4 link etc does not accomplish this. 

In that video they added a hat section, not a flat bar and welded it to the inside of the rocker and created another tube inside the rocker that will reduce the deflection/sagging but not help much with twist.

[Casgar 21]

Somewhat of a necro of this thread, but I feel my question is most appropriate here.

Anyway, I just realize that the inner rockers on 69-73 have a pretty nice slope on them where the seat sits, which makes them look way more interesting to me than before. What I would like to ask is if anyone know the dimensions shown below, and if there is any reference point on the body of the car in relation to the inner rocker? I would like to make a dummy out of wood so I can measure and see what kind of seats I can manage to squeeze in. The cross section is suppose to be the middle (A) part.

[Mach1 Driver 560]

On 4/12/2019 at 6:27 PM, Casgar said:

Somewhat of a necro of this thread, but I feel my question is most appropriate here.

Anyway, I just realize that the inner rockers on 69-73 have a pretty nice slope on them where the seat sits, which makes them look way more interesting to me than before. What I would like to ask is if anyone know the dimensions shown below, and if there is any reference point on the body of the car in relation to the inner rocker? I would like to make a dummy out of wood so I can measure and see what kind of seats I can manage to squeeze in. The cross section is suppose to be the middle (A) part.

about the only thing I can tell you is that guys have welded the front to the front plate on the torque box, but I'm not sure which end is front. That may be the passenger side, as I think I see the seat belt weldnut just above the left side of the B. You might try asking the manufacturer.