Viscosity Test Temperature

At first sight, the slope of a straight line between those two points is -0.63[3 recurring] viscosity units per degree centigrade. However, viscosity isn't a straight line function. If it were, then the liquid would boil at 115.8degC, which it probably doesn't.

What units are those viscosity numbers by the way?

There is not really a formula, because different fluids behave very differently in terms of viscosity vs. temperature. You will find charts of viscosity change vs temperature for common liquids, especially hydrocarbons, in books such as Camerons Hydraulic Data. I don't know if such info is available online. But if you've been in touch with both manufacturers, they should be able to provide you with visc.vs.temp curves or charts.

A "controlled experiment" includes holding at least something constant....

Viscosity is resistance to shear. Depends on the materials involved.

What temperature is the fluid when in service. That should be the test temperature.

What are the viscosity numbers of fresh, uncontaminated oil? What values are within specs, or values you can live with?

What is the service life of a fresh fill of oil? How long has the oil been in service now.

You said "First sample was tested at 40C the second was tested at 100C, hence my problem." Why is this the problem? Were these the temp extremes that were selected to test at?

The fluid in question is a cutting oil. New, it has a viscosity of 17.4cSt@40c. The original viscosity test was 44.1cSt@40c. When I sent out the second sample I specified to have it tested for viscosity. The lab assumed, and tested it 100c. The problem is that we had to slow down the feed to stop burn marks forming on the teeth of the gear being ground. We are looking at $7000.00 to replace the cutting oil.

This really doesn't make sense. Is it a coolant or a lubricant?

Did you have to "slow down the feed rate" after the oil was contaminated?

How many meetings will it take to equal the $7,000.00 cost of replacement of the oil, which is the only option you have if you are a competent, trustworthy shop.

Your inability to have two tests performed using the same parameters doesn't say much for your technical capabilities.

Change the oil!

I agree 100% that the oil should be changed. Unfortunately I don't have the authority or hold the checkbook. I do have a plant manager who loves to micro-manage. As to your question about the fluid in question, without giving away free publicity, yes it is. It is a special oil/coolant used specifically in gear grinders. As an oil it doesn't have a high lubricating properties as it does in regards to its cooling properties-to reduce the the burns it could leave on the teeth. I do apologize for my last post, it was venting. Thanks for the help and information that was put forth.

It si only a rough help. You may pinpoint the values obtained and draw parallels to the existing lines and see values at other temperatures. It will give you ONLY ane stimation but it is better than nothing. You can coppy the picture and print it for further processing.

If you perform the test for at least 10 points, then you can elaborate from the graph the pertinent formula and that should give you the expected characteristics of the oil.

But remember it would be for the oil under static conditions, since viscosity is affected by temperature, heat disipitation, flow characteristics, so the test maybe should be done with varying flows, temperatures, and pressures so that the true profile of the oil can be discovered.

Or call the oil vendor and ask them the formulation and characteristics of the oil under varying conditions.

I think what you need to find is the Viscosity Index which is the ration between the two temps where the viscosity is measured. This will help you determine the changes from one lube to the other.

I can't help but to comment on the information given in this post. To me, nothing quite makes much sense to me. It appears this fluid is a metalworking used to lubricate and cool the gear, perhaps sprayed on and captured in a sump and returned to the storage tank for reuse. But, not at all sure about this. As one post mentioned, simple, change the oil as it is probably contaminated, but the person writing the post doesn't say what is the contaminant. The testing data I guess showed a viscosity increase of about 14(virgin) to 44 @ 40C. If this is correct, then the lube oil is clearly out of spec's, probably due to oxidation and contamination. Of course, the oil will test with a lower viscosity at 100C than 40C., either new or used. Since we don't don't know the specs of the virgin lube oil, we don't know what type of lube oil he is using, the temperatures, and other necessary information, impossible to help very much

This is probably not much help to you.

James

If i understand the problem correctly, why is it not possible to re-test the second sample's viscosity at the correct temperature, i.e. 40c ?

What impresses me is the way commentators despise the OP question. The guy asked for a possibility to check if the 2 results are coherent and this was all. He did not ask for counsels he did not ask for any thing else but we are a lot more clever and we are able to tell him that what he did is wrong although we do not understand the question. To make it clear for some of us when grinding a gear the contact zone between steel and grinding wheel is strongly loaded from the thermal point of view. In order to avoid structural changes in the hardened steel temperature has to be maintained under a level and in order to do it a flow of oil is sprayed over the zone. It takes heat from the gear by forced convection. Now what happened is that due to a higher viscosity the boundary layer increased and thus the heat transfer was diminished. This lead to the speed reduction since heat power is proportional to grinding speed given a pressure between wheel and flank. If temperature is too high color appear on the heated zone due to chemical interactions and generation of thin layer which due to diffraction generate colors. It is clear that the oil for productivity reasons has to be changed but this was not what he asked for.

Agree with you totally.

However, for entirely my knowledge i would like to understand why sample's viscosity cannot be measured at a different temperature again. I'm not trying to make a point or anything..i just want to understand!! (incase i'm missing something)

You are correct. The question asked, I believe, is there any correlation between the two tests. The answer is no. You test sample A @ 40 & 100C and test sample B same method.

Oils have about same behaviour versus temperature so that it is possible to see if values are or not compatible by using a viscosity-temperature graph as I suggested.

The first was not for a temp range broad enough so that the following one can show if within usual error limits the 2 results are possible to compare:

Points A, B & C correspond to the indicated values. The lines passing through have same slope as the near by running lines for other oils.

One sees that :

- the values 44cStk at 40 C° and 6 cStk at 100°C are almost on same line or that if a straight line is defined by those 2 points its slope will be near to the other oils in the same viscosity range. This means that the results are comparable within the limits of normal errors due to the fact that there are 2 samples and two meausrements at different times.

- Point "C" indicates where the initial value is located and how far it is from the last values.

There are several comments to be made:

- the oil was, possibly, not from the start the best choice for the working temperature and it changed its composition

- the filtering was not so efficient as it should have been so that -I presume- metal and wheel particles stayed in the oil and participated at the chemical changes

- it is not known how long the oil was in use may be so long that it is normal to obtain such modifications

- the change can be justified by the impact on productivity and it would be better to analyse with the supplier which kind of oil has to be used in order to obtain the most economical result = maintain the low viscosity for good heat transfer for a long time under the grinding conditions.

- filter has to be, possibly, changed for a finer filtering.

This should have been the answer at least according to my humble opinion + the explanations I gave in the previous mail.

With respect to the question if a new sample has or not to be analysed the answer depends on the possibility to correlate or not the 2 results. If correlation not possible then a new analysis is compulsory but as it appears this is not the case.