Oil is taking forever to warm up after oil cooler install

[savage217]

I just installed a vpd oil cooler kit with a 180 degree thermostat. After the install I noticed oil temps took at least twice as long to get to 180 degrees. I thought the point of the thermostat I installed was to make the warm up process about the same as stock. I have bumped an old thread in the forced induction section as well as talked to a few people with oil coolers prior to making this thread. After listening to the comments it seems the answer is still unclear. Some say it should warm up as if the cooler isnt there, which I thought it would and is the purpose of the thermostat. Others say my long warm ups are normal.

It probabaly took 20 solid minds to get to 180 degrees. Before it would take 10 mins or so. I have heard people covering their oil coolers for street use but I thought that was the purpose of the thermostat? Can anybody tell me if I may have it plumbed wrong or if something doesnt sound right? here are a few pics of the setup. It may or may not help.

[m3ltw98]

You need to tighten those straight AN fittings off the oil filter cap. The gap from the blue "nipple" to the red receptacle is supposed to be about a fingernail thick.

[savage217]

Originally Posted by m3ltw98

You need to tighten those straight AN fittings off the oil filter cap. The gap from the blue "nipple" to the red receptacle is supposed to be about a fingernail thick.

Yea I found out about that a little earlier. I will make sure I tighten any of the lines that arent a fingernail or closer. Obv that isnt the "problem" but I will def fix that.

[bmwpowere36m3]

To me that seems "normal", even if the ambient is ~60°F+ it takes over 20 minutes for my oil to reach ~ 180°F (reading from oil pan). It takes a solid 10 minutes just for my water temp to get to 180°F in these current temps (~50°F).

One thing I've noticed is that my oil temp reading from the ODB reads hotter than the oil pan... till the point they both reach 180°F. Then they read the same from there in out.

[skierman64]

How much extra oil is needed for an oil fill now? Remember the engine now has to heat an extra quart or two of oil which will take longer.

I assume you're comparing apples to apples with warm up times. Same route, same speeds and similiar ambient temps.

[osborni]

Take care with the routing of the SS braided lines. They are VERY abrasive to rubber parts and body paint. You can get nylon braided tube at any industrial hydraulics (think fork lift service, or industrial equipment service in the yellow pages or similar) store or most farm supply stores.

Ideally, the hose should be held fast to the body structure with some sort of rubber or nylon damper for longer runs. If the hose is free, try to get it to route so that it will not rub or contact anything.

You might want to think about the routing of the hoses as well. You can probably take out a foot from the feed lines from the cap.

Personally, I'm not a fan of mounting the cooler so the lines enter from the bottom. Yes, I know lots of the people do it that way. It is harder for the oil flow to push air out of the system. Ideally, the hoses should enter from the top or the side so air will naturally want to rise out of the cooler.

I have not timed my oil temp rise rate, but it is just getting up to temp after a 25 minute commute to work. I have the S54 oil filter block with internal thermostat and a 625 series Setrab cooler. My temp sender is after the thermostat on the way to the cooler, so it's not terribly accurate when the oil is cool, but it is obvious when the thermostat starts to open.

[Ranger1]

I replaced my (E30) OEM oil cooler with a significantly larger one last Winter. As part of that project I put in oil temp sensors before it and after it because I was curious as to how much oil cooling would occur. As a result I've been watching my oil temps pretty closely all season. One of the many interesting lessons that came out of this was an awareness of just how slowly oil temps rise. 20min of city driving doesn't get the oil temp up to equilibrium doesn't surprise me at all. 10min surprises me.

I don't have an oil thermostat.

Oil doesnt exchange heat all that well. So it doesn't absorb it from the motor well and doesn't dump it in the oil cooler well. Therefore it's slow to heat up, and will always be a weak secondary source of engine cooling.

[Def]

Originally Posted by Ranger1

I replaced my (E30) OEM oil cooler with a significantly larger one last Winter. As part of that project I put in oil temp sensors before it and after it because I was curious as to how much oil cooling would occur. As a result I've been watching my oil temps pretty closely all season. One of the many interesting lessons that came out of this was an awareness of just how slowly oil temps rise. 20min of city driving doesn't get the oil temp up to equilibrium doesn't surprise me at all. 10min surprises me.

I don't have an oil thermostat.

Oil doesnt exchange heat all that well. So it doesn't absorb it from the motor well and doesn't dump it in the oil cooler well. Therefore it's slow to heat up, and will always be a weak secondary source of engine cooling.

Oil exchanges heat just fine. Its specific heat capacity might not be as high as water, but it can still be a VERY large percentage of the total cooling in a motor when you have a small air-oil cooler.

Just putting a small oil cooler on my Nissan dropped oil temps by 50 deg F+ and water temps by 15 deg F on the track.

[savage217]

Originally Posted by osborni

Take care with the routing of the SS braided lines. They are VERY abrasive to rubber parts and body paint. You can get nylon braided tube at any industrial hydraulics (think fork lift service, or industrial equipment service in the yellow pages or similar) store or most farm supply stores.

Ideally, the hose should be held fast to the body structure with some sort of rubber or nylon damper for longer runs. If the hose is free, try to get it to route so that it will not rub or contact anything.

You might want to think about the routing of the hoses as well. You can probably take out a foot from the feed lines from the cap.

Personally, I'm not a fan of mounting the cooler so the lines enter from the bottom. Yes, I know lots of the people do it that way. It is harder for the oil flow to push air out of the system. Ideally, the hoses should enter from the top or the side so air will naturally want to rise out of the cooler.

I have not timed my oil temp rise rate, but it is just getting up to temp after a 25 minute commute to work. I have the S54 oil filter block with internal thermostat and a 625 series Setrab cooler. My temp sender is after the thermostat on the way to the cooler, so it's not terribly accurate when the oil is cool, but it is obvious when the thermostat starts to open.

I was planning on getting some nylon or something similar for the hoses in some areas. I dont really have room to route the lines any other way. These are -10 lines so they are pretty damn big. If I was using -8 im sure it would probabaly be a lot easier and shorter. Also because of the thermostat I had to obviously lenghten the entire system. As for the oil cooler mounted with the oil flowing into the bottom, I was curious as to which way it should be mounted. Most with the vpd kit mounted it the way I did but some people with different coolers mounted it 180 degrees.

As far as how much extra oil I added. I just threw 8 quarts in when changing the oil. Im going to keep an eye on it over the next few days and check the level.

I guess my main question is should I keep the thermostat in? I put it in to slow reduce oil warm up time. I thought this meant it would basically make the system warm as if there were no cooler there until the oil reaches 180 degrees in which case the thermostat would open. Guess not?

[NeilM]

I think what people are saying is that the time needed for your oil to come up to temperature doesn't seem unusual. I still remember well over 10 years ago when I first installed an oil temp gauge (no cooler at that time) and was amazed to see how long that took. That knowledge actually changed my driving habits. 15-20 minutes, and longer in low ambients, is perfectly normal. However it may also be worth checking your oil t'stat. It's been a few years since I installed one like that, but as I recall it's both directional and distinguishes between the outbound (hot) and inbound (cooled) connections.

Regarding the -10 lines, they absolutely need to be supported to protect those hose ends from stress and vibration. I used large Adel clamps at the cooler end (see pic below, extreme right).

Your hose run lengths do look as if they could be optimized, although it's hard to be sure from a photo. I wonder if you would not have been better off to use 90 degree fittings on the engine side of the t'stat. That would cut out a bunch of bends and hose length both. For sleeving around the hoses I used clear vinyl tubing from the hardware store. Readily available and inexpensive. It's a bit stiff when new, but once it's been heated in use the tubing conforms itself nicely to the hose routing, and because it's soft but with heavy walls it protects adjacent components from abrasion nicely. It does have to be slid over the braided hose before the final fittings are installed.

Neil

[Ranger1]

Originally Posted by Def

Oil exchanges heat just fine. Its specific heat capacity might not be as high as water, but it can still be a VERY large percentage of the total cooling in a motor when you have a small air-oil cooler.

Just putting a small oil cooler on my Nissan dropped oil temps by 50 deg F+ and water temps by 15 deg F on the track.

Quick Internet search says:
Thermal conductivity of water 0.67
Thermal conductivity of oil (xW50) 0.15

Looks like a factor of 4.5 to me.

Caveats:
-Water wetter will increase "apparent" thermal conductivity of water by reducing it's surface tension which increase's it's effective surface area. --Anti-freeze has about the same crappy thermal characteristics as oil so anti-freeze is lousy coolant.
-Oils additive packages can impact thermal conductivity.

[NeilM]

Originally Posted by Ranger1

Quick Internet search says:
Thermal conductivity of water 0.67
Thermal conductivity of oil (xW50) 0.15

Looks like a factor of 4.5 to me.

I'm not sure that thermal conductivity is the metric you want for a fluid that's being pumped. Specific heat would be more applicable. For pure water that's 4.19 kJ/kg K. Oils seem to be in the area of 2 kJ/kg K, so it looks as if a derating factor of 2-2.5 is more like it.

Neil

[Ranger1]

Originally Posted by NeilM

I'm not sure that thermal conductivity is the metric you want for a fluid that's being pumped.

Thermal conductivity is what indicates how quickly heat flows to and from. That's what you want with a heat exchange medium. The key is that there is a time variable here.

Specific heat is a measure of the amount of heat energy something can absorb. There is no time variable here.

It's handy that oil will absorb more heat, but it's not so handy if it doesn't dump much of it in the couple of seconds that it's in the oil cooler. The fact that it will absorb so much heat energy per unit delta temp is one of the reasons it takes so long to heat it up.

[pbonsalb]

I think the question is whether warm up is slower after installing a thermostat equipped cooler. I believe there is still some bypass in the thermostat, so it may have a cooling effect. Certainly much faster than if no thermostat.

I have one of these waiting for me to install it. I should finish wiring my oil temp gauge first so I can measure before and after warm up times.

[savage217]

Hey guys I think I know what the problem is. Tonight I gave my dad a ride in the car to see if he thought it was taking abnormally long. I think we came to the conclusion that the temperature sender is not working properly. We checked the thermostat lines and both were hot which means the thermostat was open. We checked this out after we were driving for 25 straight mins at say 3500 rpm and the oil temp read 120 degrees even. It didnt move at all. We played with the wiring on the sender a bit and as I let the car idle there the oil temp seemed to move up a bit. As we started driving again the gauge went back down to 120 degrees.

Another thing we kept an eye on was oil pressure. My oil pressure at idle was 10psi or so which means the oil was relatively warm as the cold idle pressure is around 50-60psi Ive noticed. I am going to do a few tests to verify the accuracy of the sensor.

I also called derale, the company that makes the thermostat and they said their thermostats are one of the few that are 100% shut off until the temperature reaches 180 degrees. Since both lines were hot we knew the thermostat opened confirming the oil temps were at least 180 degrees.

One other thing I forgot to mention. I blocked off the oil cooler during this drive and still the gauge barely moved which further proves the culprit is probabaly a faulty temp sender.

[Def]

Originally Posted by Ranger1

Thermal conductivity is what indicates how quickly heat flows to and from. That's what you want with a heat exchange medium. The key is that there is a time variable here.

Specific heat is a measure of the amount of heat energy something can absorb. There is no time variable here.

It's handy that oil will absorb more heat, but it's not so handy if it doesn't dump much of it in the couple of seconds that it's in the oil cooler. The fact that it will absorb so much heat energy per unit delta temp is one of the reasons it takes so long to heat it up.

NeilM is correct, specific heat capacity is what you want to look at when analyzing heat transfer capabilities. For a liquid to solid interface you can get the convection coefficient very high with lots of turbulence or just adding surface area. For a given flow through a core, its performance is really going to be dictated by delta T through the core multiplied by mass flowrate.

An oil cooler obviously isn't going to transfer as much heat as a radiator(they're usually about 1/5 the size or smaller), but oil is a more than capable enough heat transfer medium.


I do lots of work with liquid cooling systems on fighter jets(a type of oil is used), so I have a little bit of experience with instrumented testing in this area.

[Ranger1]

Originally Posted by Def

NeilM is correct, specific heat capacity is what you want to look at when analyzing heat transfer capabilities. For a liquid to solid interface you can get the convection coefficient very high with lots of turbulence or just adding surface area. For a given flow through a core, its performance is really going to be dictated by delta T through the core multiplied by mass flowrate.

An oil cooler obviously isn't going to transfer as much heat as a radiator(they're usually about 1/5 the size or smaller), but oil is a more than capable enough heat transfer medium.

Ranger fails. Bummer.

Def: Thx for the correction.

[5mall5nail5]

Originally Posted by Def

I do lots of work with liquid cooling systems on fighter jets(a type of oil is used), so I have a little bit of experience with instrumented testing in this area.

Have you seen the enclosed housings for electronics on fighters and the like that uses a closed loop misting system? I was just checking this out at work the other day - bare PCBs being sprayed with a solution... wild!

[Def]

Originally Posted by 5mall5nail5

Have you seen the enclosed housings for electronics on fighters and the like that uses a closed loop misting system? I was just checking this out at work the other day - bare PCBs being sprayed with a solution... wild!

No, haven't seen that. Sounds like something that would be easy to break. Typically the design calls for Line Replaceable Units to have redundant systems, so that adds to weight.


I do know of some the more "ghetto" LRUs to come through my lab that just boil off water that must be topped off after every flight. Usually foreign countries that don't know how to cool stuff via heatsinks reliably tend to do that.

[JClark]

Originally Posted by savage217

Since both lines were hot we knew the thermostat opened confirming the oil temps were at least 180 degrees.

Or the t-stat is broken.

[savage217]

Originally Posted by JClark

Or the t-stat is broken.

One is always cool when I start it up while the other is hot. I doubt the thermostat is bad. It's possible but it's brand new. I'm gonna put the temp probe in some boiling water and see what it reads.

[NeilM]

Originally Posted by Def

I do lots of work with liquid cooling systems on fighter jets(a type of oil is used), so I have a little bit of experience with instrumented testing in this area.

Jeez Def, it's the end of the driving season and here Ranger1 and I were gearing up for a nice little internet flame war to pass the time, with all the forceful logic and imagined expertise we can come up with to support our respective points of view, and you go and jump in with actual rocket science! Or jets, close enough.

Where's the fun in that? Don't you know it's against the very ethos of the net?

Neil

[Ranger1]

Originally Posted by NeilM

Jeez Def, it's the end of the driving season and here Ranger1 and I were gearing up for a nice little internet flame war to pass the time, with all the forceful logic and imagined expertise we can come up with to support our respective points of view, and you go and jump in with actual rocket science! Or jets, close enough.

Where's the fun in that? Don't you know it's against the very ethos of the net?

Neil

It was a near thing. I was about to mention that I was a mechanical engineer with 2 semesters of thermo and 2 semesters of heat xfer, so you could just stfu. It would have been bad to say that just before Def's smackdown. <g>

That being said, I'm still struggling with this. It's fine to say that one could get the convection coef high by adding surface area, but that's kind of a tangent. We're not talking about comparing different oil coolers, we talking about oil's ability as a heat exchange medium vs. water.

I submit that the rate of heat xfer is porportional to the medium's heat xfer coef, but independent of the medium's heat capacity. The key is "rate" because the oil only spends seconds in the oil cooler.

If oil was as good a heat exchange medium as water, we wouldn't be using water.

If I'm wrong, help me understand why I'm wrong.

[NeilM]

Originally Posted by Ranger1

It was a near thing. I was about to mention that I was a mechanical engineer with 2 semesters of thermo and 2 semesters of heat xfer, so you could just stfu. It would have been bad to say that just before Def's smackdown. <g>

I'm electrical, but everyone knows that an EE starts every explanation by saying "Imagine that this copper wire is instead a pipe full of water." That way I'd claim fluid dynamics expertise. See how that works? Excellent tactic!

I submit that the rate of heat xfer is porportional to the medium's heat xfer coef, but independent of the medium's heat capacity. The key is "rate" because the oil only spends seconds in the oil cooler.

My logic is that if you had a stationary horizontal column of fluid and applied heat to one end, then the rate of heat transfer to the other end is a function of the fluid's thermal conductivity. This is a hard model to apply in real life, since there would be convection currents within the liquid, plus the question of the containment material, but hey, what's science without an idealized experiment?

But in our real world case the fluid is pumped, and pretty damn vigorously at that. We don't have to wait for heat to meander from one place to another: it soaks into the oil and gets moved along in a gross physical way, rather than at the molecular level. So we care more about how much heat a unit of cooling fluid can hold rather than how fast it can conduct that heat to the adjacent unit of fluid and so on down the line — specific heat more than thermal conductivity.

The better oil coolers have internal vanes to assure turbulence in the oil for improved heat transfer to the cooler. I've no idea how to conceptualize the engine's heat transfer into the oil, except to say that it seems to happen well enough. Modern engines do often have oil jets squirting on the piston undersides to help cool them.

If oil was as good a heat exchange medium as water, we wouldn't be using water.

But you're right, oil isn't as good.

Even looking at it my way it's less than half as effective. And even hot oil is pretty viscous, so the pumping losses are much higher with oil. Then of course there's the issue of oil degradation if it gets overheated.

But there have been oil cooled engines, or at least air/oil. Air cooled from the outside, oil helped cool from the inside. There's obviously a long line or air-cooled Porsches, but they mostly had dry sumps with high capacity external oil tanks to assist in cooling. Somehow they even managed to cool those 1000hp turbo charged 12-cylinder 917 engines. Back in the 60's Honda made an air/oil cooled 4-cylinder 1300cc passenger car engine, not to mention their unique 3 litre V8 Formula 1 unit of 1968. (Yeah, I am old enough to remember it...)

Oil seems to be pretty useful for auxiliary cooling of water cooled engines, but clearly water has carried the day for the principal cooling needs.

Neil

[Ranger1]

Originally Posted by NeilM

My logic is that if you had a stationary horizontal column of fluid and applied heat to one end, then the rate of heat transfer to the other end is a function of the fluid's thermal conductivity. This is a hard model to apply in real life, since there would be convection currents within the liquid, plus the question of the containment material, but hey, what's science without an idealized experiment?

But in our real world case the fluid is pumped, and pretty damn vigorously at that. We don't have to wait for heat to meander from one place to another: it soaks into the oil and gets moved along in a gross physical way, rather than at the molecular level. So we care more about how much heat a unit of cooling fluid can hold rather than how fast it can conduct that heat to the adjacent unit of fluid and so on down the line — specific heat more than thermal conductivity.

Continuing to beat this horse only until someone helps me understand why I wrong......

Your example is a heat sink. The issue is heat transfer. A heat sink can absorb a lot of heat. Think iron brake rotor as opposed to a (hypothetical) plastic rotor. Iron has a much higher specific heat then plastic. It takes a lot of heat energy to raise the temp of the iron rotor 1deg. But that doesn't tell us anything about the iron rotor's ability to absorb or dump a lot of heat in a hurry (thermal conductivity).

The thermal conductivity of a material can't be ignored just because fluid is pumped. If a heat exchange system dumps 1000btu/min, and you swap the coolant fluid with something that has half the thermal conductivity, you're going to get 500btu/min dumped, or you'll have to double it's flow rate.

Anti-freeze sucks as a coolant because it's thermal conductivity is almost the same as oil.

Ahh, screw all that. I looked it up. With one exception, there is no heat capacity variable in heat exchanger calculations. It's all thermal conductivity. The exception is the Dittus-Boelter correlation (no, really) which doesn't apply to oil very well (doubters look up Prantl number).