Alternative to JB Weld???

[fyrwrxz]

So far, the aardvark is still holding.....testing still in progress. Reports later.

 

[CarVac]

The force is distributed between the centering rings, it doesn't go to one joint "first," as you say. It depends on the relative stiffness of the centering rings and the motor mount tube. If the motor mount is much stiffer, then both centering rings will have the same stress. If the centering rings are much stiffer than the motor mount tube, which they may be depending on the inner and outer diameters, then indeed the stress will go to the lower one first.

 

[David]

The force is distributed between the centering rings, it doesn't go to one joint "first," as you say. It depends on the relative stiffness of the centering rings and the motor mount tube. If the motor mount is much stiffer, then both centering rings will have the same stress. If the centering rings are much stiffer than the motor mount tube, which they may be depending on the inner and outer diameters, then indeed the stress will go to the lower one first.

Nice practical clarification to the model. We still need some math.

 

[CarVac]

It's a statically indeterminate problem, so you have two simple analytical methods: force/flexibility or displacement/stiffness. I'll do the first method.

If the centering rings deflect .1 mm per Newton, say, and the motor tube between the two centering rings deflects 1 mm per Newton, then the whole system's flexibility is 1/(1/.1+1/(.1 + 1)) mm per Newton, which equals 0.0917 mm per Newton. That shows that the first centering ring takes most of the load, if they are much stiffer than the motor tube. Therefore, when 1 Newton is applied, the first centering ring handles 0.917 Newtons (since it deflects the same amount as the whole system) and the second centering ring only takes .083 Newtons.

If the centering rings deflect 1 mm per Newton, and the stiffer motor tube only deflects 1 mm per Newton, then the system's flexibility is 1/(1/1 + 1/(1 + .1)) mm per Newton, which equals 0.524 mm per Newton. They split the load approximately evenly: the first centering ring takes .524 Newtons and the second centering ring takes 0.476 Newtons.

This isn't really useful for hobbyists, since the stiffnesses of the parts aren't really known, and additionally this model does not take into account the stiffness of the body tube, or the presence of through-the-wall fins which may be glued to both the body tube and the motor tube.

 

[David]

Perhaps not useful, but this thread started with a discussion of JB Weld, and led to comparisons of epoxy strength. Just got me thinking about how strong the epoxy really needs to be. Why are we debating 23 Mpa verses 50 Mpa, for example, if 10 will do it, with good technique? Keeping the model overly simple offers the chance to look at the bond between the CRs and the tubes. Just a curiosity.

 

[CarVac]

The reason to use better epoxy is so that you can use less of it. Fewer centering rings = less weight = more performance.

 

[David]

The reason to use better epoxy is so that you can use less of it. Fewer centering rings = less weight = more performance.

So, other than experience, how do you know when your epoxy is better-enough to drop down a ring? Which gets back to what I'm asking: Is there a way to know, in the the way an engineer might work out the strength of a bolt needed for a bridge, the epoxy strength needed for a particular design? I'm not an engineer, but I doubt she says "I'll just use a better bolt and use less steel." She would probably use some numbers and the numbers would be based on something.

 

[CarVac]

There is a way to know, but it involves making lots of measurements that, while definitely possible, are just not worthwhile time investments at least in my eyes. There are too many variables to test: which epoxy, fillet size, fillers, thickness of the centering rings, tightness of fit of centering rings, stiffness of the centering ring material, stiffness of the motor tube, stiffness of the outer tube, etc.

We can't assume linearity, so for example you would have to test each epoxy with different fillers and in different fillet amounts: some will behave better with one set of parameters than another would, and vise-versa.

 

[David]

That makes a lot of sense. Ton of variables. It is possible to simulate and get an answer, but no practical reason to do so, and the answer would be a range anyway. Curiosity satisfied. Go with good epoxy, materials, technique and some experience.

 

[MarkII]

He is referring to your mixing alcohol with the epoxy reducing the strength of the epoxy, not the epoxy reducing the strength of the baffle.

Yes, but that's a given. The coating isn't applied for any structural purposes, so the bond strength isn't relevant. It is used here as a coating, not as an adhesive for the purpose of structural bonding.

 

[GregGleason]

I have been thinking through this again, with regard to the JB Weld formulation.

The MSDS for JB Weld does list iron as an ingredient. But my guess is that it is really carbon steel powder, which is typically over 95% iron, followed by up to 2% carbon, followed by a mix of other elements. You would think that at the machine shop where it was created, there would have been an abundance of steel powder available. Hence the listing of iron as a portion and not listing the minor elements found in steel.

Looking at this from another angle, perhaps it is not the thermal conductivity of iron, but the reduced thermal conductivity of steel. If that is the case, then aluminum would be a worse choice as a high-temp amendment. A better choice from a lower thermal conductivity standpoint would be 316 stainless steel, with a slightly better choice being titanium. Perhaps it is the elements that do not conduct heat well that serve it better being a high-temp amendment to epoxy. That's my theory anyway.

Perhaps some testing is in order.

Greg

 

[CarVac]

I highly doubt that conductivity or lack thereof contributes to the high temperature resistance in any way. The user testimonies claim they plug up holes in engine blocks using JB Weld, and those stay hot long enough for the entire epoxy blob to heat uniformly.

 

[GregGleason]

I'm not so sure. Given that JB Weld is general purpose epoxy with amendments, I think the first stop is to look at the amendments they use. The chief amendments in JB Weld (from what I have been able to glean from the MSDS and other sources) are iron (and in my theory, it is really carbon steel powder, from machining operations) and calcium carbonate (limestone). There is a lot of beautiful limestone in the Hill Country of central Texas, so that is a readily available source.

The key to any high-temp epoxy is a high glass transition (Tg) stage. It is perhaps the lack of conductivity of carbon steel, along with calcium carbonate that raise the effective Tg for epoxy. I'm not an epoxy chemist, but perhaps there is something that happens at the molecular level that raises the Tg. Regardless of what is going on and how it is going on, I think that the key is in the amendments.

Regardless, it is fun to kick around.

The best way is to run a controlled experiment to find out. Even if the Tg is raised 200 degrees F, that would put Aeropoxy within the upper limits of NFPA 1125's motor case requirements for external temperature (200°C (392°F) per Section 7.4.1, 2007), so safe enough to use in the motor area. But I'm not sure how the strength would be affected with the amendments. Likely an increase, but only to a point.

I do think one could run a crude experiment with a heat gun and a timer to find out. Hmmm...

Greg