How does rubber cure physically? 2

[examorph]

Hello,

I am trying to simulate the curing of rubber but I am confused with how rubber actually cures in reality. From looking at rheometer results I can see that initially the viscosity of the rubber drops as it is heated, the rubber flows then the viscosity/stiffness of the rubber starts increasing.

With this understanding I am planning on doing the following; run a heat transfer simulation in which I can see how temperature changes within the material when heat is applied to the surface and then apply conditions so that if any part of the material reaches a certain temperature its viscosity starts increasing at a certain rate in time.

Is this how rubber is cured or is my understanding of this incorrect? I just ask because with this theory for example if a small block of rubber takes 10 minutes to cure and a large block takes 60 minutes this would mean that the heat took 50 minutes to get to the center portion of the large rubber block to fully cure the entire material (from surface to center)which from my understanding is unlikely, there are also a few other things that just don't add up with this theory.

I am really getting confused with this and would appreciate any help at all with this problem, even if you can provide anything on this topic please just let me know.

Thank you.

 

[btrueblood]

You have the basic idea. Yes, it's a time/temperature relationship, with a typical Arrhenius rate equation that depends on the materials used and type of cure, etc. The reaction is not linear with time, but a decaying exponential function (google Arrhenius rate equation or look at

http://en.wikipedia.org/wiki/Arrhenius_equation

- note that you need to look at the entry for "reaction rate constant" also to see the effect of reactant depletion), i.e. as cure agent and the available sites on the base polymer are depleted in the chemical reaction (the "concentration" of the precursor chemicals if you will), the rate of reaction slows. So, "full cure" is something of a misnomer, as the absolute depletion of cure agent never occurs in theory, just asympotically approaches it. I think 90% cure is a typical endpoint used in rheometer tests...but don't quote me.

"I just ask because with this theory for example if a small block of rubber takes 10 minutes to cure and a large block takes 60 minutes this would mean that the heat took 50 minutes to get to the center portion of the large rubber block to fully cure the entire material (from surface to center)which from my understanding is unlikely, there are also a few other things that just don't add up with this theory."

Rubber is an insulator, i.e. a poor conductor of heat, so yes, it takes more time to cure a thicker/larger section of material when the heat is applied at the outside surface. Uncured rubber, without all of the cross-links, is a poorer conductor than cured rubber, so this affects the rate equation as well. Then, factor in the uncertainties (did the cure agent get properly dispersed in the milling/mixing steps? Are the temperature indicators on the press heater controls reading accurately?).... and most places probably over-cure by a factor of 1.5 to 2x...but better that than an undercured part.

Interested what you think doesn't add up with the theory. It's understandable to me that you are confused, the "science" of rubber is a black art, and a lot of information that is helpful seems to be tied up in trade secrets.

 

[btrueblood]

Hmm, re-reading your post, this line caught my eye:

". From looking at rheometer results I can see that initially the viscosity of the rubber drops as it is heated, the rubber flows then the viscosity/stiffness of the rubber starts increasing"

It is not intuitive, well wasn't to me, but a rheometer curve is a somewhat ridiculous standard if your goal is to determine the cure time of a molded part. In rheometer testing, the temperature to drive the curing is generated by the internal self-heating of the material as it is continously mechanically mixed/stirred by the rotating disk in the mechanism. This, plus the thorough, continuous mixing of the materials, mean that a rheometer test is going to give a much faster cure rate than the actual molding process (with conduction heating only, and no continuous mixing) would ever be able to achieve.

 

[examorph]

Thank you so much for your help, it really has helped me understand the curing process better especially the part about cured rubber being more of a conductor than uncured rubber, this should be fairly straight forward to add to the simulation.

The thing I mentioned that did not add up with the theory was that when I have seen people working with rubber, to see if the part is cured they simply remove the die top and indent the outer surface of the component, if the rubber has its elastic properties in which it rises again back to its original position it would be an indication that the component has cured. This made me think that rubber cures in a homogeneous was in which everything is cured together which goes against the theory that I learnt however, after reading your post I am now thinking that this may have been an inefficient way of testing if the rubber component had cured.

One last thing, the rheometer being used is a moving die rheometer (MDR) and from my understanding (which is very little) this does not mix the material therefore it does not generate heat within the test piece and the majority of it is transferred through heat conduction. If this is not true could you please suggest an alternative method of obtaining results on how a component cures which are fairly accurate.

Again, thank you for your help I really appreciate it, as you said now days everything is a trade secret within the rubber industry and finding information is very difficult.

 

[GrahamBennett]

"moving die rheometer (MDR) and from my understanding (which is very little) this does not mix the material therefore it does not generate heat within the test piece and the majority of it is transferred through heat conduction"

This is true. The die only moves by ±0.5° so there no shear heating is generated.

 

[Compositepro]

Examorph, Rubber World magazine is an excellent source of information on rubber technology, although you cannot bypass studying basic chemistry.

http://digital.ipcprintservices.com/publication/?m=9911&l=1

 

[examorph]

Thank you for the confirmation and also the link, it was a very interesting read.

I now have a new question which is also cure related and will help with my simulation. I was thinking about how components are molded, o-rings for example. If you have two o-rings both the same cross section one a much larger diameter than the other, theoretically from my understanding of how rubber cures they should both cure exactly at the same rate as the cross section of the material cures the same and the same temperature is applied to both however, from seeing o-rings get made in reality the larger o-ring requires about 15 minutes to cure whereas the smaller diameter o-ring only requires 5 minutes. This makes no sense, and has really confused me because it goes against my current understanding of how rubber cures.

I would really appreciate if anyone could answer why this happens. Thank you.

 

[GrahamBennett]

If the O-rings have the same cross sectional area and differ only in outside/inside diameters, then it is definitely possible to cure them in the same time at the same temperature. In fact, you can lay the mould out so that the cavity for the smaller o/d O-ring lies inside the one for the larger o/d O-ring.

 

[tom1953]

For the larger diameter o-rings, is the mold much bigger, heavier, greater mass, and needs more re-heating to get back to curing temperature after the previous o-rings are removed after cure? As Graham indicates, the same cross-section o-rings should cure in the same time, regardless of the diameter (all other things being equal, which is rarely the case . . .)

Tom Jablonowski, TSE Industries, Inc.

http://www.tse-industries.com

 

[RubberDog]

Adding to Tom1953's message: when you are looking for the o-rings being molded (large o-ring, small o-ring / same cross section), you definitly need to consider the heat transfer and masses.
If the press plate is the same, let's sayd 20x20inches, you probably can fit just one o-ring of 18 inches... all the inner portion will be empty.
If you have 002's o-ring, you will have all the plate full of o-rings, so much more mass to heat, so you need more time to the press stabilize again during the molding cycle.

Also, the preparion method (heat history) will impact the curing... do the o-rings have the same heat history? Was one done by calandering and the other by preforming?

And last, but not least, without trying to discourage you, I have seems people trying to simulate the curing, but it is always using more information than the rheometer can give you.
It has always the thermal coeficients added (thermal conductibility, activation energy, etc). The rheometer probably can approach only of the shear that a compression press do (and if molded slow).
By experience I have seem same compound, pure injection being molded in 2minutes. If injection/compression time = 1min, due the excessive shear that the molten compound suffers)...


RubberDog