Series vs. parallel oil cooling
 

I know that running oil coolers in series is better for max cooling but results in a higher pressure drop. Is this that big of a problem with the rotary engines?

Application is various nonturbo 13B engines (I'll be realistic here :( ) meaning small pump and large pump style engines, short (40-90 sec) run times at low speeds so most of the airflow comes from the cooling fan. I currently have a GSL-SE oil cooler mounted in its stock location with some ducting to prevent air from going around it.

My goal is to be able to keep the engine cool in real time. Currently, I've been finishing runs with coolant and oil temps in the 230 degree F range. I can get coolant down to ~190 before the next run but oil is more difficult, so I want to focus on oil cooling. Plus, sometimes a repeat run has to be made right away, and some regions split into smaller run groups so there IS no cooldown time. I'd like to be able to take my car to those events again.

I have an FC cooler that I plan to mount above the existing one. Series and parallel would be about equally as difficult to plumb, obviously parallel would cost a lot more, but $100 more for fittings is still cheaper than another killed engine.

Not overly familiar with the early series engines but pressure drop is usually only a couple PSI over the whole cooler so a second one in series should be fine. My guess is that in your case just adding a second cooler will fix up the problem.

I like the idea of parallel better, but I notice that the fd and rx8 are in series, so maybe it doesn't matter.

The only thing is the pre fd coolers are long, vs the later ones being tall, so the old ones should have more pressure drop

i know the second gen oil coolers have a build in termostat, so in theory when the oil is cold and thick, it bypasses the cooler and goes right back to the engine. and when the oil is hot a thinner it is less resistant to getting pushed through the cooler.

in reality the you'll see high pressures when it's cold, due to it's resistance to being pushed into small gaps like the bearings. so as long as there are thermostarts i don't think it will matter either. IE if it's hot in the first cooler, in series to the second cooler, it'll bypass and and still get to the engine with the same pressure.

have you modded your oil pump passage? I've see it result it a much faster oil pressure response and 10 to 15 PSI increase.

Parallel will give you more cooling, not series. The cooler is more efficient when the temperature differential from the air to the oil is higher.

There is a bunch of easy ways to run parallel without spending a ton on AN fittings. I usually run a remote filter on the race cars so I buy a mount that has two in, two out and use one in and two out. Through the coolers. Coming back into the engine I like to bring both circuits back to the stock return location. I started to make my own custom fittings that use a stock banjo bolt with the head drill through and then weld an AN fitting from the hydraulic store welded on to the top.

One thing that I always wondered about was if one cooler was cooling the oil more than the other and bringing the oil back to two locations with a big difference in oil temps.(Like to the front plate) It probably wouldn't be an issue but bringing the oil back to one spot blends it

What I'd been reading is that parallel will have less resistance to flow, but series will result in better cooling in marginal applications. If one cooler gets more airflow than the other, the cooler oil in it will have more resistance to flow and the oil will preferentially go through the hotter one.

I can't poke holes in the theory like the "water moving too fast" myth for cooling systems, plus there's enough anecdotal evidence to support it.

Ultimately, two would be better than one no matter what, but if one way would result in oil volume and pressure problems, then I'll avoid it. Most setups run plate style oil coolers which would seem to me have less resistance to flow than what Mazda used.

I'm starting to have second thoughts, though, and am now considering trying to rearrange the existing coolers in what I think is called a V mount configuration.

jgrewe nailed it.

Personal experience is this - Customer's ITS road race car came in a couple years ago with a series setup. Comes in for engine management install and tuning. We cover that part. Car makes great power but oil pressure is lower than I'd like to see. We're seeing only about 70 psi WOT 3k+ RPM. Ideally, I'd want to see 100+. Tell owner this, they run it anyway, with no changes. They had bought the car already built as it sits, and, I think, convinced there is not an issue.

I'm asked to drive the car in a couple regionals that year and the car performs very well. However, oil and water temps are fairly high (230-ish), and oil pressure during extended running is no more than 70 psi. I think perhaps the engine has a stock FC pressure regulator in it, and recommend they have the regulators swapped out for Mazdaspeed or REW regulators.

Week after I win a couple races with it, the owners call me. The engine is locked up. They bring it down to me and we tear it down together. Front main is seized, noticeable wear on all other bearings. Oil pump is broken. One of the rotors is in 4 pieces. It's a TII pump FWIW. I build them a new engine with the typical ITS setup. New TII pump, REW regulator, Mazdaspeed open e-shaft jets, race bearings, etc.

Start the engine in the car with the old series setup and immediately have lower than expected oil pressure. Pressure on first start (cold oil) at idle is around 70 psi. We should be seeing 100+. As the oil temp comes up, the pressure goes down. To around 25 psi at idle. We're still seeing a max of 70 psi with warm oil at 3k+.

I'm thinking one or both of the coolers is clogged. I send them out for ultrasonic cleaning. No change. I bypass one of them and run the engine with just one cooler. With one cooler we have 100+ on cold start, comes down to around 60 at idle, and 100-110 warm at 3k+. Good! So the other cooler is bad right? Nope. Plumb that one in by itself and the results are repeated. With just a single cooler, either one of them, the pressure is good. Plumb them in series and we loose 30+ psi at 3k+ with warm oil!

One other thing we found while experimenting was that one of the coolers still had the thermostat bypass while the other had it removed and the bypass blocked. With relatively cold oil (the working thermostat bypassing), we would see good oil pressure at 3k+. Once the oil warmed, and the thermostat closed, the pressure at 3k+ dropped.

Working backwards, I think we answered the question of "how the hell did the oil pump break"? Without a pressure gauge on the line at the front cover outlet I can't say for sure, but I would have to guess that with the huge restriction obviously imposed by the series setup, that we would be bumping the pressure setting of the front regulator and working the heck out of the oil pump to do so.

^ There is a seemingly-obvious difference between your customer's series-plumbed oil cooler setup and a set of FD or Rx-8 oil coolers, plumbed in series from the factory. Some guy rigged up those dual coolers in his garage whereas the OEM FD and Rx-8 coolers were actually engineered to be run in series. So the oil cooler thermostats, the piping, the OEM oil pump, the dimensions of the coolers, the ducting, etc all work together as a system. The OEM dual coolers may not have the highest oil cooling capacity on paper but at least they were track-tested by real experts.

Most turbo piston engines have the equivalent of a single FD oil cooler or use coolant to cool the oil like some of the 1st gens did. So in that comparative sense the OEM dual coolers are still high capacity.

C.Ludwig with that said, pressure drop in the system is dependant on both plumbing and the cooler used. Ive run 3 different cooler setups, all in series, and never had pressure issues. I log oil pressure pre cooler and have the a gauge post cooler.

From an engineering standpoint more flow and greater temp differentials both improve the efficiency of the cooler. series will maximize flow, while parallel will maximize temp differentials. Series will have greater pressure loss over the coolers and parallel will have greater pressure loss over the plumbing. IMO its all a trade off, properly setup both are well proven setups.

...That's exactly the real-world data I needed to hear, C. Thanks a bunch!

Summit sells -10 Y-blocks for $35 a pop and -10 straight fittings are cheap enough. I'll use the Summit blocks but I'm not thrilled with their hose ends. Fine for the low-pressure fuel applications where I've used them, but I'll stick with Earl's for oil.

Yes, I'm dithering. :) It all depends on how low and forward I'd be able to cram the radiator in so it would fit at an angle enough that I could mount an oil cooler next to it. As it is, my radiator fan to water pump clearance is small enough that the fan rubs from time to time.

Of course, if I could un-marry myself from the idea of using stock oil coolers, I could just run one big square one under each headlight. But then we're rapidly exiting the notions of "using stuff on hand". Plus I'd need an oil thermostat, and I don't think the GM ones that I've scavenged would have enough flow.

Any rules against a system for spraying water on your oil cooler during your runs? Just throwing it out there.

None at all. The rules structure is very lenient. Basically, if it is recognizeable as an RX-7 (*) and meets some very rudimentary safety minimums, it's good to go. Engine and drivetrain are 100% free, as is suspension.

I've considered a water sprayer, but my concern with that is that the days where temperature control is a serious problem tend to be dusty. Dust blows off okay, mud tends to stick around.

http://sphotos.ak.fbcdn.net/hphotos-..._1220060_n.jpg

(*) The rule that bothers me. I could lop a foot off of the nose of the car SO easily, it's all empty space out there in front of the hood mount's crossbeam, but then it could be argued that it is "not recognizeable" anymore.

the FD and Rx8 coolers are taller cores with shorter flow tubes = less pressure drop per cooler?

Parallel will provide as much or more cooling capacity as series coolers. Because the oil spends more time in each cooler being the flow is divided between two rather than flowing through two in serries at the same flow rate as if it were only one cooler. Plus, both coolers are operating at the same high inlet temp rather than one having less temperature differential (Thus, less heat rejection).

Pressure drop is far less through two coolers in parallel as the flow is roughly cut in half through both coolers. Pressure drop in a closed loop system is solely based on flowrate.

Thus parallel piping (good parallel piping with Y-fittings or reverse return arangement) is better in every way. And fairly convenient.

Not to say series piping arrangements are bad. But the very primary reason they were employed on the FD and RX-8 was due to cost. MUCH cheaper to run them in series when they are at opposite front corners of the car.

FCs have an advantage in having all the connections on the same size. thus it is just some extra costs in fittings to make them work in parallel.

PS: I am a mechanical engineer who specializes in heat transfer, you can listen to me with faith. ;)

http://iscracing.net/images/0oilcooler.jpg



Not sure how well it works, but ISC is usually pretty on top of it.
www.iscracing.net

I considered that one as well. It looks like a huge two pass cooler. Probably very effective and pretty cheap to plumb.

I ultimately decided to go with two stockers so i could have the thermostat capability without spending a bunch of money on an aftermarket external thermostat. And i had two recently reconditioned stock coolers already in hand.

peejay - another thought might be to add a fan or two to your cooler(s). The relatively low speeds of rallycross doesn't provide much in the way of air being forced across the coolers you do have. Just a thought...

ISC has some issues with that cooler and I think they are going to change the design. They are having pressure problems that lead to cracking, I have a friend that has one and its cracked twice. Inlet side pressure is stupid high.

that works. back in the day a friend of mine had a FMIC, and 240 degree oil temps, cause the IC blocks the oil cooler.

he put 2-3 computer fans on it, and viola, oil temps were more like 190-200

I have an electric fan for this purpose. The oil cooler is ducted to the radiator so that the airflow can't bypass the oil cooler to get to the fan. Fan is hardwired to the ignition system - if the key is on, the fan is on.

Current procedure is to finish the run, get back to grid as quickly as possible, shut down engine as soon as possible, and turn ignition back on to get the fan turned on. Fan then runs until the car is cool or the next run is due. This cools the engine a lot more quickly than leaving it run, surprisingly enough. Sometimes I start and run the engine a few seconds at a time to cycle the fluids around, but it seems to be unnecessary.

I was thinking about the computer fan thing, since the 120mm jobs seem to grow on trees around here, but I was worried about stifling airflow.

I'm also thinking of attacking the cooling problem from the other direction - a small additional coolant radiator mounted in the heater hose loop. If nothing else, it would allow me to run without the heater on :)

Never tried fans, but I really pay attention to oil cooling. Coming from an air/oil cooled Porsche background, I have seen some pretty extreme oil cooler setups.

Oil cooler in 911 Turbo Race car, its fairly large and does its job perfectly-
http://i218.photobucket.com/albums/c...r/IMG_3615.jpg

http://i218.photobucket.com/albums/c...r/IMG_3617.jpg

http://i218.photobucket.com/albums/c...r/IMG_3619.jpg

Little oil cooler on Eclipse GSX track car-

http://i218.photobucket.com/albums/c...r/IMG_3620.jpg

I have had no real issues running oil coolers in series, but I can see how others have.

Love it! That's a nice cooler.

...and looking at it, the Porsche cooler looks a lot like a bar and plate cooler in design with large tanks and many short tubes, while the Mazda coolers force everything to go through two or three very long tubes... in each direction.

So, I can see how Mazda coolers in series can cause restriction and not the other ones.

I'm not impressed by engineering degrees, I'm impressed by real-world evaluations. What works on paper never seems to pan out in the real world...

The oil cooler in the 911 is not OEM obviously, but very effective.

I read everywhere about how the OEM cooler was all you ever needed, blah blah blah, but I could not bring myself to use it. Instead I used a single "stacked plate" oil cooler similar to the coolers that used to be made by "Long" and now are sold by B&M etc.. I used a 11x12-ish cooler (would have to go actually measure it) and never had anywhere near high oil temps even at altitude and continuos high RPM. Obviously using e-85 affected this, but I think too much is made of the 2nd gen coolers abilities. I could be wrong, but this has been my experience.

I assume you've gone back to talking about the RX-7 setup? Interesting.

The setup that I'd most like to emulate is on this rally car (may look familiar to some). I forgot where I found this pic, but it's probably 7-8 years old, but if you look through the screen you can see the radiator and oil cooler are side by side.

BTW - I can probably cover the stock oil cooler with new 120mm fans for $30. Experience with computer electronics says that the fans do not tolerate any amount of contamination with anything. Dust, moisture, anything...

Yes sorry, on the FC a stacked plate style oil cooler is used. I may use two of them if needed when I revert back to pump gas, but between the ducting and the efficiency I have no real worries about my oil cooling.

ive seen a few cars side by side, i like it. one they are both in front, two, you don't have the oil cooler hanging out where it gets hit by curbs etc.

a lot of the 60's mercedes sedans are like this, and yes it has an FC sized oil cooler for the engine.

toyota had a turbo celica at the historics one year, and it had the radiator next to the IC.

While it is true that testing is an important part of engineering, doing things on paper first saves a lot of hit and miss screwing around. Obviously you have not run into any good engineers. Oh, and your entire car from the factory was done on paper (by degreed engineers) before ever even going into it's first prototype. Then guess what, it was tested and redesigned, by....you guessed it.....engineers.

But if it makes you feel better, own a road race FC that we run enduros with. So i think i know enough about oil cooling in the "Real World" to be part of the discussion.

Further, half of earning an engineering degree is lab time where the "on-paper" is directly related to physical tests and results. Thus creating EXPERIENCE as part of mathematical capability. When an degreed engineer looks at a fairly straighforward problem like this they can anticipate the relative results WITHOUT doing anything on paper, and consistently be correct. Because, not only have i done the calculations thousands of times but i have seen how the calculations ACTUALLY correspond with "Real World" results...

I have, but I've also run into plenty of "Ivory Tower" engineers who don't seem to want to believe that their on-paper theory might not be taking all factors into account.

Damn it, another project for my rotary-dominatrix: parallel the Mocal oil coolers. Thanks, no really, thanks alot! :burn:

Good discussion, and definitely topical for me.

I'm running Paul Ober's old track car. NA FC, stock internals. It came with dual FC oil coolers plumbed in series. From what I recall, the thermostats have been removed. One of the coolers is in the stock position while other sits behind the nose (like Dan Lemon's old E-Prod car) and has its own ducting cut into the bumper region.

Here's the problem - I have *too* much cooling. In a summer HPDE at VIR (paddock temps in the mid 90's), my oil temps (measured at the pedestal) rarely exceed 160F. I have to tape over one of the nose openings in order to get things up to 180 (see picture).

So before I would get into comparisons about serial v. parallel in terms of cooling efficiency, I would instead look at the ducting first. It seems that even serial setups with enough ducting can be more than sufficient (at least for my car).

The oil pressure discussion is probably more important here. My pressures are slightly lower than I would like (definitely not the 100psi that Chris mentions) so I am considering gong to a parallel setup this year and see if that has any significant impact.

-b

Honestly , I think that dual coolers is overkill and that it might need a better location or fan, or both. I trashed the stock sytem on my FC and redid it with a Taurus fan that was including the oil cooler in the airflow. And I have a hard time getting enough temp, even in 98 deg days.

https://www.rx7club.com/picture.php?...ictureid=27802

Lot of air and a good cololer with the T-stat replaced with a plug.

GD

Well i never said it was not crazy to run a cooler without a stat!

No air IS the problem... and I wouldn't consider running without a thermostat! Even with a thermostat, I only see oil temps of around 140 degrees in the pan when tooling down the highway. And then there's the issue with cold days... if there's no salt on the roads, I'm driving the RX-7, even if it's still only 40 degrees out. Under rallycross conditions, 40 degree days are the only ones where I don't have much overtemp problems.

You also have an FC engine bay, meaning you don't have an engine 1/4" away from your fan to block the airflow. I didn't have an oil temp gauge before relocating the hood latch, but coolant temps dropped from ~190-200 on the highway to ~170. Amazing what helping the air out of the engine bay will do for radiator efficiency. Didn't help the rallycross temps all that much.

I did a little shopping and with the cheapest fittings I could find (not sure I like that) it's still going to run me about $180 just in plumbing to add the second oil cooler. A couple 10AN-18mm adapters, five straight -10, one 90 degree -10, two Ys, and a few feet of hose. Adds up quick, doesn't it? Sad thing is, the only pieces I could conceivably make (18mm adapters) are the cheapest individual components at $7 per so it doesn't even make sense to do that.

edit - I just realized that I don't need six ends, I need EIGHT. D'oh, that makes it worse!

Have you thought of just keeping the nose of the car pointed into the wind?

Has anyone ever thought of welding two oil coolers together? what would be the cons of that?

Ha! While that is the fast way around the course, I think you may have helped me stumble upon a justfication for my Tim Allen Syndrome:

I have to make the car go faster so it will run cooler. :lol:

I know I can out-drag modified Evos on the straights, the only problem is that the pesky course designers rarely install any straights :)

wow that looks pretty good actually. It stays cool with that little ass radiator. really.....

those bastards!

the oil cooler setup i arrived at with mine seems to work really well.

the car came with an FC cooler shoved in the nose, they welded tabs on the round cross bar and just let the cooler hang out. since this is ugly, and i intend to park the car, and probably put it into the weeds too, i wanted the cooler out of the way.

after many hours of trying to get the FC cooler and hoses to fit, i think i found the perfect solution.

i used a gsl-se cooler, came core size, different hoses and mountings. the cooler is actually bolted to the radiator panels, and sits horizontally behind the round cross bar. i added a little duct, actually right now its a cereal box, to go between the undertray and the cooler, and it works great. plus its totally out of the way.

Exactly what we found without the blowing up engines part. I run 2 stock FC coolers in parallel and both with the thermostats still in them. I can get 120+ psi cold, but around 80 - 90 psi hot. The temps barely crepp over 200 degrees and most of the time, even during slower tracks with tons of rpm and low airflow we barely creep over 210. At Mosport, even in the summer, I have to tape over part of the grill to get oil temps above 165 degrees. The car does 3 hour enduros with no oil pressure or temp issues. Even after heat soaks during pit stops.

I strongly recommend 2 FC coolers in parallel. It is cheap and effective.

Eric

@ Barry, I don't want to quote that big pic again. At first read about the oil only really being hot for the first three inches fo that cooler it sounded like it made sense. But after thinking about it for a second it seems to me that side tank is just acting like an extension of the hose that is bringing the oil. The oil that travels to the far end of the tank before it runs through the core would only lose temp like it was moving through an aluminum tube. No?

It would be very interesting to hit various parts of the cooler with an IR thermometer to see what it says.

This was my plan for about 6 months now, i guess ill just have to R&D it.

I'm going to use straights wherever possible but a 90 will be required at the FC cooler's top fitting, and there is already one at the engine's inlet. They're radiused ends and not right-angle, so I don't see it causing a problem, in light of the way the rest of the oiling system is shaped.

10psi/1000rpm is fine for a small block Chevy, which is the engine Smokey Yunick invented that little idea for.

Although really, if I'm seeing upwards of 10k at times, I'm not running enough pressure by that logic. :)

I'm going to attempt running Mobil 5W30 this year instead of Valvoline 20W50. I think I've solved my oil pressure issues with the RX-8 front cover gasket. (It just plain works)

Have you considered blocking the front cover port and coming straight out of the side of the front plate?

PS- looks like you are having great fun with the rallycross.

Barry

Decent Quality AN fittings are certainly not cheap. My cost for the parts to make the lines was well of $300. I dont think it is any cheaper to do dual coolers vs one big one unless you allready have two FC coolers that are clean and in good condition. Cleaning for me was 125 per cooler. I had to buy a total of 3 because one had an unrepairable leak...

I have, but realistically it won't solve anything since I want that front regulator there protecting my oil lines and cooler, so that junction is going to be there no matter what.

It's one of those back burner "that'd be cool" things to have, like a proper loop line arrangement, but the reality is that I need to keep things simple so I don't have to engineer new parts every time I have a failure of some type. Rerouting the oil arrangement is low on the ease/benefit scale.

I have tried several different oil cooler configurations on my 13B PP road racer, and for the last several years I have had very good luck with two stock coolers plumbed in series. I get pretty decent airflow at road racing speeds. One cooler was not quite enough, so I went to two coolers. Per race engineer expert Carroll Smith's writings, I plumbed them in parallel, and my oil was still on the hot side. I finally took my InfraRed Pyrometer and checked the temps, and one was hot and the other was cool! I then replumbed them in series, and it has kept my oil cool for several years of road racing.

I just started co-driving a friends Toyota truck in Rallycross, so I now know the format of your runs. I think one cooler with electric fans would be the way to go. I know the computer fans are the right dimension, but I don't think they would flow enough CFM of air for you. What about a fan from the car's heater system? don't they have about a 4" outlet size? They have much higher CFM than a computer fan.

Nice seeing you on the forums again.

Your experiences are part of where I got the idea that parallel wasn't as good as series for really getting the temps down.

The sickness spreads. :) I remember some guys from around your parts came up here in 2009 with their green 2002.

I was poking around at the car and I noticed that while I had the top of the oil cooler well ducted, the bottom had a large area where air could flow right under it, so I'm going to pay closer attention to that first.

I'm also going to (as soon as I can round up another person to push my car into the shop!) see about fitting a Neon radiator in the nose. It's a single row radiator, pin-mounted, and the automatic models had an option for two very beefy large fans. It will require an FC water outlet and some water bleeds, since the radiator has no provision for a cap. But, I also can lay the radiator down so that air can get OUT of it.

Sounds like you had a restriction or blockage in one side of your parallel system so, in effect, you were still only running a single cooler.

As a mechanical engineer with experience in fluid flow and heat transfer I maintain a properly configured and installed parallel system will out-perform an in-series system.