Shrink fit "frictionized" by application of a colloidal silica solution?

[Grunchy]

I'm having a hard time finding information anywhere in the literature. Co-workers are drawing a blank as well.
The issue we have, on an existing assembly, is a relatively heavy shrink fit between a gear and an intermediate shaft (width 2.00"; ID of gear Ø3.1880/3.1890, 63 microinch finish; OD of shaft Ø3.1930/3.1940, 32 microinch finish). The shrink fit asks for +250°F on the gear, -100°F on the shaft.
So far so good, but the spec then asks for Nalco 2329 colloidal silica to be applied to both surfaces prior to assembly.
A vendor is asking for alternatives because of long lead time of this one colloidal silica solution; but I'm at a loss of what the performance spec is of this existing specification, let alone how to judge a bunch of alternatives.
What I've read about it so far is that one potential use of colloidal silica is as a "frictionizer" for paper products. It doesn't exactly make sand-paper, but it adds a frictional roughness for a grippy feel. I cannot imagine for what other purpose we would be specifying the colloidal silica for.
I suppose the distinction between different colloidal silica preparations is how much friction each adds, but so far none of the vendors are able to provide that kind of information. This has to be the distinction, I can't imagine what else it would be.
Thanks Much for your advice.

 

[SwinnyGG]

That's a lot of interference.

Does the math say you need this stuff?

With an interference fit that heavy, I'm having a hard time imagining that adding anything that isn't an adhesive between the parts is actually having a net positive effect on the power transmission possible through that interface.

 

[EdStainless]

It almost sounds like it being used as an assembly lubricant, that won't reduce the grip.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[btrueblood]

The only thing I can think of, that hasn't been mentioned yet, is it possibly a means of reducing galling, allowing the parts to be disassembled at a later date?

 

[Grunchy]

Hi guys I did some math... used the tribology-abc calculator:

http://www.tribology-abc.com/calculators/e6_2.htm

Inputs:
di=0 (solid shaft)
d=80.9752mm (3.188")
do=314.6806mm (12.389" root diameter)
L=50.8mm (2.00" engagement)
δ/d=0.001882 (0.006" interference ÷ 3.188" diameter) <-- highest interference
µ=0.12 (modest friction factor)
Ei=210GPa (AISI 8620 shaft)
Eo=210GPa (AISI 8620 gear)
vi=0.3 (Poisson's ratio 8620)
vo=0.3 (Poisson's ratio 8620)

Solving I get interference pressure 184 MPa and transmission torque 11585 N·m (8544 lbf·ft).
The final deliver torque is 35,000 lbf·ft and the final ratio is 16:74, so this interference fit needs to transmit 35,000 x 16/74 = 7568 lbf·ft.
So safety factor is 8544÷7568 = 1.129x or about 13%.

First of all - calc check ok?
Second - I assume the purpose of the colloidal silica is to increase the friction factor inside the shrink fit, but I've never heard of it being used this way before. (Unfortunately - the guy who had designed it has since left the company, otherwise I would just ask him directly.)
So I'm hoping somebody here can confirm this usage of colloidal silica? Obviously 13% safety margin is pretty thin. (with max. interference - would be inadequate for min. interference.)

Incidentally, the 2nd part of that calculator verifies the hot/cold temps for the shrink fit. The only number I could find for thermal expansion coefficient of AISI 8620 is 44 x 10-6 / °F which is about 79.2 x 10-6 / K.
Inputs were:

δ/d=0.001882
ai=79.2 (10-6 / K)
ao=79.2 (10-6 / K)

Gave me -23.8K for the shaft, +23.8K for the gear. Total temp differential 47.6K = 86°F. So actual temp differential of 350°F might provide enough clearance for a clearance fit + space for the colloidal silica coatings. But that's assuming the average thermal expansion coefficient is actually 44 x 10-6 / °F. I saw other numbers for 4140 more like 23 x 10-6 / °F.
I'm not too concerned about this aspect, we've been manufacturing this assembly for years now & no issues - so I'm assuming it must work ok or we would have heard about it.

Thanks For Any Advice & Insight

 

[tbuelna]

Grunchy,

Did you check to see if that Nalco 2329 compound can be applied to a surface at -100degF?

As an alternative, you might consider using an anaerobic adhesive compound (ie. Loctite) designed for assembly at elevated temps. Using Loctite in addition to your interference fit will increase the torque capability of the connection quite a bit. Consider that the actual contact between the mating surface asperity tips of an interference fit only amounts to around 20% of the total area. Applying anaerobic adhesive fills in the gaps and creates an adhesive bond in the remaining 80% of area not in contact. There are several Loctite retaining compounds that have a cured shear strength of >2500psi. And since your shrink fit interface has a fairly large surface area (~20 sq.in.), a quick calculation will give you an idea of how much added torque capability Loctite compounds can provide.

 

[Grunchy]

Good point tbuelna & EdStainless, but I think the surfaces must be coated & dried prior to heating the gear & chilling the shaft. It's like a suspension of colloidal silica in a water carrier, the water would boil on the hot side and freeze on the cold side. So it can't possibly be used as an assembly lubricant... ?

Thanks for the suggestion of the Loctite retaining compound, I think that makes better sense. I'll definitely look into it as an alternative.

TTYL