First, I want to make clear this is about structural foam. I think several of the replies reference expanding insulation foam, like the kind you get from the hardware store. That's like comparing the tin foil in your kitchen to structural steel. Sure, they're both metal, but the similarities pretty much stop there. And you couldn't have a relevant discussion about making, say, subframe connectors out of tin foil vs structural steel. Same with insulating foam and structural foam. They're both foam, but that's about the only thing they have in common and there's no relevant comparison for chassis stiffening.
Second, I absolutely read the articles. I actually went back and re-read them. Just because I mostly disagree with your conclusions about their relevance for use in A-bodies, ie, guys adding structural foam to their 40 + year old cars in their garages, doesn't mean I didn't read the articles. There are several VERY important processes involved in the articles that aren't being replicated by just injecting foam into an A-body in someone's garage, and a few relevant points that are not addressed at all.
All of the articles used finite element models for structural analysis. Not only did they use FEA models, they picked the application points for the foam based on the results of those FEA models. That's not the same as just picking the joints and corners of a chassis and adding structural foam. Those joints were chosen based on the results of FEA analysis, making the use of those joints specific to the chassis that was analyzed. Considerations were made based on the shape and size of the cavities as well. I don't see any FEA analysis going on of the A-body chassis to determine where the structural foam should be added. Not only that, but the fact that FEA models existed for the chassis' in these articles means that the engineers would have had the relevant information on cavity volumes to determine the amount of foam needed in each joint/cavity. Not to mention the location of all the relevant baffles and internal stiffening plates etc that would confine the foam. The GM article even points out that the modal testing didn't match the FEA testing because of the injection sites used and internal baffling changing the way that the foam filled the cavities compared to what was modeled. The Chrysler article very clearly states that their models assumed that "the material expands to form a block of foam perfectly bonded to the sheet metal." I don't think you can assume that with a chassis that's been out in the world for well over 40 years. The chassis' tested in the GM article were current production year models that were pulled off the assembly line. Brand new cars. The chassis used in the Chrysler article isn't specified beyond it being an SUV and the graphics appearing to be of a Jeep Cherokee or similar, but given that they already had the FEA models I would guess that it was either a brand new, or at least very new, chassis.
I also don't see how you're going to constrain the foam to the joints in any of the areas you've circled. How will you guarantee the foam won't run along the entire length of the inside of the rocker, and expand to only partially fill it? You'd have to add baffles or internal structures. Not to mention that the door supports and bases of the a-pillar structure where they attach to the rockers are not open to the rockers, there's no corner shaped void to fill. You'd have to fill the rocker section and upright sections separately. Same with the A-pillars. They're hollow. You wouldn't be able to add foam to just the top corner, you'd have to fill the entire pillar unless you baffled it. And again, same for the cowl section you circled at the base of the A-pillar. That whole cowl is an open space, you'd have to baffle it if you wanted to keep the foam near the base of the A-pillar. Otherwise it would just run to the bottom of the cowl. Doing that would involve removing the upper cowl completely, which is not something a lot of us need, or want, to undertake.
Your point about the foam "locking" into the joints isn't valid. Look at the Mazda paper again. Very little bending strength was added in the partially filled cavities. If the foam wasn't bonded fully to the metal, you only got the bending strength of the foam and metal separately, and the foam has almost no bending strength by itself. That's why they made such a big deal out of the adhesion strength of the foam.
As far as the US Cartool and XV testing, comparing that to FEA and modal testing by major automotive and engineering companies is a pretty big stretch. Don't get me wrong, they make good products and some of the testing methodologies used are relevant. But you can't forget that both companies sell a product, and there are videos of the testing for visual impact. Which may certainly have changed the methodology and procedures used. It's like basing a chemical engineering discussion on the properties of OxyClean using one of their commercials. Does that mean the tests shown in the commercials aren't valid? Not necessarily. But the point of the commercial was not to conduct a 100% valid scientific test, it was to have a visual impact to sell product. Same with the videos comparing chassis stiffness. Not enough of the methodology or process is discussed to use the video as the basis for structural engineering analysis.
I'm not saying that the structural foam is useless, or can't be used for this application. But you only have a couple papers on structural foam used in modern cars that are heavily entwined with FEA. That's not what you're doing. There's no information on how any of it is going to work on a 40+ year old car with rust, corrosion and dirt with the foam applied by a novice in a garage based on picking important looking joints and guessing at the volume of foam needed. And there's no articles included about doing a comparison between the structural foam and say, adding some gusset plates. That's mentioned in the GM article, but it's mentioned along with how big a deal it is to retool an assembly line to do stuff like that. Adding structural foam is a better deal for an assembly line making tens of thousands of cars. Adding it to an A-body is a one off. They can do the research, FEA, and testing to back up the FEA, that's what they do. And they're going to already have all the product and tools to add the foam. That's easier for them. In a home garage adding a few heavy gauge sheet metal gussets is a heck of a lot easier, and you don't have to be an expert on structural foam application to do it. You just need a MIG welder and some metal, and most of us are already set up to do that. Sure, if we were mass producing that would be more costly and time consuming, but not for a one off. And I seriously doubt any of us are going to do enough testing with the proper methodology to produce a valid engineering analysis to prove what works better in the end.
I mean, knock yourself out. Seriously, I'd love to see how you go about doing it- prepping the chassis, constraining the joints to add the foam, the end product and even the anecdotal "seat of the pants" results that you get. But I don't think you're going to prove that adding the structural foam is better than say, adding subframe connectors and torque boxes. And adding those is a lot easier for most folks at home in their garage than what's going to be necessary to ensure a quality result with the foam. But that's just my opinion.
