What is the basis for a max. oil temp of 194F?

[jimhokie]

I frequently see a maximum allowed oil temperature of 194F (90C) in machine bearing applications. I know the oil life decreases rapidly at higher temperature due to oxidation, but what is the basis of the 194F limit? So many hours of life?

Also, in a ball bearing, is there any known correlation between the oil sump temperature and the peak temperature of the oil in direct contact with the rolling elements, i.e., localized hotspot as the ball shears through the oil film?

Any insight will be appreciated. Thanks...

Jim

 

[Skogsgurra]

90 C is a nice and round number just below 100 C. There is no dramatic change when going to 91 or 92. But, given tolerances in measuring equipment, I have always thought it was kind of a "convenience limit" on the safe side. Like many speed limits on the roads seem to be.
Anyone else knows if there is a better reason for 90 C? What about synthetic oils? Same limit?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Tmoose]

Just like so many things in life, many factors sometimes each can tug on the string

Different cage materials have different operating temp limits.

http://www.bardenbearings.com/PDF CD/CAGESBRO.PDF

Bearing steels remain dimensionally stable only up to a particular temperature. Special processing must be used to raise the dim-stab temperature.

Typically both effects would laugh at 200F, so the theoretical lube life shortening would appear to be a main driver

 

[romke]

there may be two reasons or this. the first reason may be viscosity: most oils specified are in the ISO VG system and manuactures may feel that using a particular grade, the actual viscosity at 90 C is about the minimum the bearing can cope with. the other reason is oxidation: a rule of thumb says that from that point on every 10 deg C more doubles the rate of oxidation, resulting in oilthickening and the formation of deposits that may harm the bearing. whena using better baseoils (API group III, IV or V) 90 deg C does not necessarily have to be the limit, but a certain minimum viscosity is needed to get acceptable bearing life.

 

[electricpete]

So it was metnioned we have several limits and we try to keep below the least of them:
bearing dimensional stability - problems around 250F
lubricant aging
cage temperature limit
viscosity variation with temperautre

The last of these is worth talking about. As temperautre increases, viscosity decreases. We had sliding thrust/radial bearing with manufacgtuer specified temperature limit of 100C (alarm at 90C). When installed the machine was running very near the limits (which is a story in itself). Part of our strategy for addressing the condition was understanding the basis for the limit and investigating whether it could be increased. The machine OEM told us we could go up to 115C if we used synethetic oil with higher viscosity index so that the drop in viscosity with temperature woudl not be so steep. In this particular case viscosity was the limiting factor.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Eratosthenes]

Bearing components are typically tempered at 325 degF, dimensional stability shouldn't be an issue below that temperature. For operation above that, the tempering temperature must be adjusted and hot-hardness can affect bearing capacity depending on the bearing steel used.
Bearing life greatly depends on the oil film thickness as a function of operating viscosity, speed, and load. A good relative film thickness lambda can increase bearing life 4 to 5 times. Selecting the correct oil helps a lot.

 

[electricpete]

250F is a number often cited for max temperature to heat a bearing for installation (for example SKF Bearing Installation/Maintenance Handbook). Also Machinery's Handbook states: "To facilitate the assembly of a bearing on a shaft it may become necessary to expand the
inner ring by heating. This should be done in clean oil or in a temperature-controlled furnace at a temperature of between 200 and 250F. The utmost care must be used to make sure that the temperature does not exceed 250F. as overheating will tend to reduce the hardness of the rings. Prelubricated bearings should not be mounted by this method."

I am pretty sure I have seen somewhere that the bearing manfuacturers recommend to consult an application specialist for applications where 250F is exceeded (of course temperature is only one of many factors considered in applciation).

Looking in a FAG document I saw a mention of 150C as limiting temperature before bearing steel would be affected. That is closer to 325F than 250F.

I suspect there are many varieties of bearing materials/treatments have no problems well beyond 250F, but 250F is my number unless I get something else from the bearing OEM.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[jimhokie]

Thanks for all the responses and thoughts. We're OK with the bearing material temp. limit...it's just the oil I'm concerned with. The bearings of concern are a duplex pair, size 7315D running at 3600 rpm. The only time oil temp could become a concern is if the ambient temp approaches the maximum of 122F, which would be an unusual condition and could push the oil temperature past 194F, but for specification and testing purposes, we need to cover this base. We are also locked into the oil we use: 2190 TEP per MIL-L-17331. I understand how the oxidation rate approximately doubles for every 10C increase in temperature, but still not sure of the basis for 194F...so many hours of life at this temp? Is there some rapid drop-off in viscosity past this point? Does anyone know how rapidly the viscosity decreases past this temp, or is this effect highly a function of the specific oil used?

I'm trying to write a specification requirement for allowed oil temperature rise above ambient that will take all these factors into effect, and that can be used to test a pump at a less than maximum ambient temperature to demonstrate that each unit will not have excessive bearing temperature rise characteristics, which would push the oil temp past 194F at the maximum design ambient, and also taking into account a margin for unit-to-unit variations (maybe about 10F max?).

 

[crjones]

On the SKF website there is a oil viscosity calculator that would allow you to see their choice and how yours matches up.

 

[electricpete]

See "Diagram 5" here:

http://www.skf.com/portal/skf/home/products?maincatalogue=1&lang=en&newlink=1_0_119

7315 Bearing has Dm = (75 + 160)/2 = 117.5
Your speed is 3600 rpm


On Diagram 5, find 117.5mm on horizontal axis, read up to 3600rpm line (between 3000 and 5000), read to the left to find about 8 cSt minimum viscosity at operating temperature.

http://www.matweb.com/search/datasheet.aspx?matguid=148fbf4d0536447da976d86b639432e3&ckck=1

2190TEP at 100C has viscosity of 9.7 cSt. So the oil should allow going up to and slightly beyond 100C without going below SKF's minimum viscosity recommendation.



=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Tmoose]

Are these the same units that were the stars of the "Thrust bearing overheating" thread last year?

 

[jimhokie]

crjones and electricpete...thanks for the tip on the SKF viscosity info. That fills in a blank I needed.

Tmoose...yes, this is same pump! After much testing and correcting of a few design flaws that dropped the operating temperature by 20-30 degrees, we are still left with a bearing that just generates too much heat to meet our unnecessarily restrictive spec requirement, thus I'm now revising the requirement to something more reasonable yet adequate.

The biggest issue turned out to be a result of the specified housing material (CA6NM) having a lower coeff. of thermal expansion than the bearing outer race, thus as the bearing heated up, it rapidly outgrew the tight design clearance in the bore. Opening up the bore slightly corrected this problem.