Dry Sump Research

[FlashArts]

I am currently doing some research on dry sump systems. More than likely it will be fitted to a gxr600, but I would like to to know what other teams feel about the topic, since I am new to engineering and FSAE.

So far I know there is two routes you can take
1) getting a pan/system off the shelf
2) designing pan/system specific for car setup.

Obviously the second option would be better engineering wise, but I am concerned with it being reliable, and the first option would be simpler, but I am concerned with cost and possible weight gain (since design might be overkill for FSAE).

I am looking for pros and cons along with helpful advice from teams who have had success with their drysump systems

[rmk36]

We tried it. It certainly helped keep oil pressures where we wanted them, but we had to run electric water pumps because the wet sump pump, and mechanical water pump were driven off the same shaft, and the dry sump pump was in the way of the mechanical water pump.

This created charging issues, cooling issues, etc.

It also was expensive and added weight.

I'm not saying there is nothing to gain from a dry sump. I feel there are a lot of advantages of it, but we didn't have enough time to really develop it on our first go at it.

[MechanicalChris]

Interesting..

rmk36, can you pinpoint why you had cooling issues with the electric cooling system?

FlashArts, I would suggest looking into your own system (pan, pump, etc). Unless there is a manufacturer who makes specific dry sump pans for our applications.. I may be wrong but I daresay they don't exist and would probably need a whole bunch of modifying anyway.

[FlashArts]

I fear exactly what rmk36 explained. Our team is definitely looking into lowering cg, but the weight gain and increase complexity might do more harm than good. I read with the drysump you get a power increase due to less windage, but I wonder it's enough to counter the extra weight.

And yea please explain the issues you had with cooling. I read somewhere else that one team used the stock wetsump pump as the pressure pump, and just added two other scavenge pumps along with a pressure release valve.

I agree Mechanical chris, does your team run a drysump?

[Rotary Sprocket]

[QUOTE]Originally posted by MechanicalChris:
Interesting..

rmk36, can you pinpoint why you had cooling issues with the electric cooling system?

The cooling issues arose because the electric water pump we were using couldn't produce enough pressure to get the water through the small nooks and crannies of our kawi motor. We tested the flow rate of the electric pump by itself and then the flow rate of the water through the engine block. When the pump was forced to pump through the engine block the flowrate was cut by almost 66%. We ended up having to run 2 pumps in series to get the necessary pressure for the system to operate efficiently.

[ESaenz]

Originally posted by Rotary Sprocket:
rmk36, can you pinpoint why you had cooling issues with the electric cooling system?

The cooling issues arose because the electric water pump we were using couldn't produce enough pressure to get the water through the small nooks and crannies of our kawi motor. We tested the flow rate of the electric pump by itself and then the flow rate of the water through the engine block. When the pump was forced to pump through the engine block the flowrate was cut by almost 66%. We ended up having to run 2 pumps in series to get the necessary pressure for the system to operate efficiently.

Which pumps are you guys using?

[rmk36]

Originally posted by ESaenz:

Which pumps are you guys using?

I think it was a Davies Craig EWP.

Ryan

[Macros]

Those pumps have been used numerous times in setups with dry sumps.

[MechanicalChris]

Indeed, the pumps aren't by any means "rare" amongst FSAE teams, I don't mean to state the obvious but I would look into other components of your system.

I know that we have run an EWP for many years now, haven't really had an issue (apart from obvious issues).

At this stage we are not dry sump, the team has previously done some testing with it and actually ran the system, I am not exactly sure why it was no longer used. I think it was due to the fact that it became overly complicated with the turbo system as well. Being aspirated though should provide simplicity for oil lines/pump use.

Yes, there is engine efficiency gains to be had by running a dry sump system, also some ring packaging advantages if you are going that far to run a different set of rings, which add to the efficiency gain on their own.

[Jan_Dressler]

It is not really that difficult to design your own dry sump. At UAS Karlsruhe, we use ours since 2008. Be a little conserative concerning the flow rate of the scavange pumps and hose sizes, place your pickups right, design your oil tank right, and you won't have any problems. You can always try to make it lighter after it works properly...

Cooling: We used the Davis Craig EWP80 for the last 3 years without any problems. In fact, we will have a much smaller pump this year...
Cocerning electric water pumps: The efficiency might not be as good as with a mechanical one, but you get better cooling in the lower rev range and better control of the cooling flow rate in general.

Power gains: I don't think the dry sump gives you that much power. Remember that it needs power to drive the scavange pumps.

[rmk36]

A lot of teams run the EWP with no problem, just make sure you have a strong enough charging system, and a motor without too much restriction in the cooling passages.

[Jimmy01]

Wow, I'm surprised so many people have had no troubles with the electric water pumps. We have about 5 broken ones lying around, they pretty much all leak (although new design is better). An electric pump failure has also been the cause of atleast one event failure for us. There are some bosch pumps around which are much more reliable.

Dry sump tank design is really important, make sure you read up about them and look at pro designs. Sufficient air volume above the oil is crucial and make sure the breather can flow enough so the tank doesn't get pressurised.

[Rex Chan]

Dear original poster,

Although I agree with many of the above posts, I'm going to put down my experience with dry sumps on 4 cylinder motorbike engines for the record. We had issues with the dry sump in 2009, but everything got sorted in 2010, and we now have a quite reliable oil engine system, by FSAE standards anyway, I think.

We run a CBR600RR 03-06, with custom dry sump pan which was designed in 2005. The engine guys this year have designed and are about to start making a new "flat" sump pan, based on the UWA design.

We ran a 2-stage (6L/min/1000RPM) Pace Products scavenge pump (steel gear rotor) in 2009, off the mechanical water pump shaft. In 2010, due to a CAD stuff up, we didn't have enough room to fit that pump anymore, so bought a 2-stage (8L/min/1000RPM)Dailey Engineering pump (alum roots blower). Both worked fine, based on oil pressure logging (VDO mechanical sensor on 2009 car; SSI peizoresistive on 2010 car).

We ran Davies Craig EWP80 (7.5A) in 2009, and EWP115 (10A) in 2010. Our radiators were single pass downflow in both years, of the same core size: 470mm X 312mm X 40mm deep. The EWPs were controlled by the EWP controller, with set point at 85C. Thermofans kicked in at 91/89C. Due to the large radiators, we were always able to keep the engine temps under control. However, we never tested under hotter than low 30Cs day, so maybe at 40-45C summer day (possible maybe a handful of days in an Australian summer), our car would overheat.

EWP leaks: although I didn't notice leaks with the 2009 car, the 2010 car with EWP115 does leak sometimes, usually when the engine is running and the coolant water is hot. Over a day of testing (maybe 2 hours of driving), we lose about 50-100mL of fluid. However, our pump is hard mounted onto the chassis, when the instruction recommend soft mounts. One issue we had with the pump is the screws used to hold the two halves together; I drilled the holes out and used M3 bolts to make sure nothing came apart, as I don't trust screws on a vibrating FSAE car. This may have caused the leaks.

As another poster pointed out, you will not get anywhere near the 115L/min free flow figure. We actually did a bucket test a few months ago, and got about 30L/min.

Windage was not a reason for us to continue using a dry sump: I think car engines have crank shafts low enough to touch the sump oil, but the crank in a motorbike engine is quite out of the way. Anyway, I do not think FSAE is an engine/power competition: my saying was that all the engine has to do is push the car around. Most 4 cylinders will make more power than the singles, but as long as the engine works, it will probably have enough power.

Dry sump oil tank breathers: make it a simple vertical tube in the top of the tank. We run a -16 hose, but its probably way too big. The reason is that in 2009, they tried to be clever and have an upward facing hole in the tank (to only allow air to escape), but so much oil is flying around inside, that oil went into the breather anyway. The BIG problem was that the upward facing hole created a kind of one way valve, which only let oil exit the system. This led to a lot of overflowing catch cans and loss of oil out of the oil system. Don't make our mistake. Thus, a big ass hose lets any upwards flying oil fall back down very easily.

Alternator/Charging/Battery issues: the main electric loads on our system are injectors, CDI ignition box, fuel pump (6A), EWP (10A), and thermo fan (10A). Normally, the fan is off, and battery volts is around 13.5-14V (charging). When the fan kicks in though, battery volts drops to 12.5V, which we interpreted as overloading the (stock) alternator. So, you can either find a way to fit a bigger alternator, use a bigger battery, or reduce current. It sup to you.

Weight/complexity of dry sump. I was lucky and got handed a working dry sump system, so to me, they're not very complex. I do agree there are more parts than a wet sump, but I worry about a wet sumps oil supply issues. You can run an Accusump, but that adds weight+parts, which may only be used in an emergency; so would it not be better to have a system that is more robust to start with? Also, a good sump pan will let you lower the heaviest part of the car (a 55kg engine) down; (although, Missouri ran a high wet sump, maybe 75mm high, and had no apparent CoG issues).

My main pros: good oil supply, and lower engine CoG.

The main cons: packaging all the parts, cost of scavenge pumps, Speedflow/AN hose + fittings

Tricky things: getting a pressure relief valve into the system, and feeding the internal pressure pump. Basically, anything to do with the insides of the engine has to very precise.

Advice: make sure the catch can is easily accessible, and is easy to empty, because you will doing a lot of both...

Now a question for you all: how does a dry sump affect the "wet" clutch in these engines? I heard there are oil passages in the clutch basket, which lubricates the discs, then drips down, but have not actually checked the diagrams/engine internals.

PS. OP, feel free to send me emails for any further questions, although the fourum lets the knowledge stick around.

[Jimmy01]

If you are getting a lot of overflow oil, then your tank design is the culprit. It is hard to fit a good tank on an Fsae car, i would suggest running about 2 liters oil (tank level not total system volume) in a 5L tank. Yes, this will be hard to package! A baffle above the oil inlet is also a good idea, as is a breather back to the engine (into head cover is the best option). -12 back to the overflow bottle would be sufficient.

Clutch: yes all(newish ones anyway) motorbikes have oil feeds to the clutch baskets/plates but some are inefficient and get away with it because of all the oil sloshing around it. If you have clutch problems that you suspect are caused by lack of oil/cooling then some simple mods can usually fix it.

[FlashArts]

I think I have had enough information to aid in my decision, thanks to all that participated.

[Jan_Dressler]

Originally posted by Rex Chan:
We ran a 2-stage (6L/min/1000RPM) Pace Products scavenge pump (steel gear rotor) in 2009, off the mechanical water pump shaft. In 2010, due to a CAD stuff up, we didn't have enough room to fit that pump anymore, so bought a 2-stage (8L/min/1000RPM)Dailey Engineering pump (alum roots blower).
Are you sure? We use Dailey Engineering pumps since 2008 (very satisfied with them), but I'm quite sure that Bill Dailey won't offer you something with less than 16l/min/1000rpm per stage. That would be about 4x the flow rate of your pressure pump (considering a 2-stage scavange pump), which will be a very safe system (and will get you a considerably amount of crankcase vacuum even without special cylinder rings).

Originally posted by Rex Chan: We ran Davies Craig EWP80 (7.5A) in 2009, and EWP115 (10A) in 2010.
Why on earth do you need that?

Originally posted by Rex Chan: Dry sump oil tank breathers: make it a simple vertical tube in the top of the tank. We run a -16 hose, but its probably way too big.
Right, it is too big. And why use a dash-something hose if you can use the lightest, cheapest hose you can find somewhere?

Originally posted by Rex Chan: My main pros: good oil supply, and lower engine CoG.

The main cons: packaging all the parts, cost of scavenge pumps, Speedflow/AN hose + fittings

Tricky things: getting a pressure relief valve into the system, and feeding the internal pressure pump. Basically, anything to do with the insides of the engine has to very precise.

Tricky? Not really... Just use the stock PRV and place it somewhere in your new oil pan, in a way that it acts as usual - feeding the oil back to the pressure pump inlet.

[Kirby]

Tricky? Not really... Just use the stock PRV and place it somewhere in your new oil pan, in a way that it acts as usual - feeding the oil back to the pressure pump inlet.

I'd be surprised if any teams do separate to this unless they run one of the pace designs with the integral pressure stage and adjustable regulator.

I think Rex was more commenting on (and I'd tend to agree with him) when designing your dry sump pan to bolt to the engine, you have to be rather accurate in any sort of connections that engage internal passages or pickups. PRV location, Internal Pressure Pump Inlet etc.

Especially tricky with the R6's as the internal pump pickup is inclined on a 15degree incline to the sump flange, so something that is able to engage that and line up on the dowels...is a challenge.

We tossed our engine upside down on a CMM and built a fairly accurate model of the internal pickups, which formed the basis for cast dry sump pan design.

Fun times..

[murpia]

Originally posted by Rex Chan:
Alternator/Charging/Battery issues: the main electric loads on our system are injectors, CDI ignition box, fuel pump (6A), EWP (10A), and thermo fan (10A). Normally, the fan is off, and battery volts is around 13.5-14V (charging). When the fan kicks in though, battery volts drops to 12.5V, which we interpreted as overloading the (stock) alternator. So, you can either find a way to fit a bigger alternator, use a bigger battery, or reduce current. It sup to you.
Not exactly on topic, but these days there's no real excuse for a negative charge balance when kit like this is so cheap:

http://uk.farnell.com/tenma/72...frequency/dp/1283640 (http://uk.farnell.com/tenma/72-7224/clamp-meter-with-frequency/dp/1283640)

Just make sure you choose one with DC current capability.

Regards, Ian

[Rex Chan]

A 15 degree angle does seem very tricky - its perfectly horizontal on a CBR600RR, and in line with the bottom of the engine block (with wet sump off).

With regards to the PRV, we had (2009/2010) the actual valve inside the pan, but the oil flowing back into the oil tank, near the bottom. A few years ago, they dumped the excess oil back into the dry sump, but the scavenge pumps couldn't keep up. So, OP, try to avoid that if you can.

I agree that -16 kevlar braided hose is overkill for a breather, but we used it for a couple of reasons: we had some lying around, and it was part of the oil system (so we used known oil + temp resistant hose, because not all hoses can take 115C). The silliest thing was that we secured it via hoseclamps at the oil tank end.

About the specs on the Dailey scavenge pump: I could be wrong on the specs. I just remembered that the 3-stage (actually only used 2) Pace pump was 6L/min/1000RPM, while the bigger 2-stage Pace was 10L/min/1000RPM, and that the Dailey was somewhere inbetween. What I can be sure of is that it was the one from this page, for pre-07 CBR600RR: http://www.daileyengineering.com/motorcycle.htm

We run electric water pumps for the same reason most FSAE teams who run dry sumps run them: the stock water pump shaft is being used by the scavenge pump. Jan: by the way your question was asked, I'll assume you think we have too big a water pump. The electric pumps don't like backpressure, and their flowrate falls off dramatically. We found a UNSW/ADFA paper that measured the flow of the Suzuki GSX-R600 mechanical water pump, which showed flow rates up to 65L/min @ 9000RPM, which is close to our average RPM operating range for our CBR600s. This would mean an FSAE car running a MWP could flow around 65L/min. However, the EWP does not pump anywhere near that amount of water, due to the backpressure in our system. So they don't even match the stock pump; and are not close to being oversized. As a result, we run a much larger than stock radiator.

Admittedly, we have an aim engine temp of only 85C (due to the EWP controller we run), which is lower than the stock thermostat of fully open 90C (btw, we have removed the thermostat, otherwise the restriction would stop water flow completely).

This year, we will be controlling the EWP115 with a 200Hz PWM out of an M400 and a solid state relay (SSR) powered by a MoTeC PDM. We can thus try a higher ET aim temp. The radiator was also reduced slightly in size, so we'll see how that goes.

[Rex Chan]

murpia: funny you should post that - I've got that exact multimeter! (off ebay or Deal Extreme though)

Anyway, how does that help with the low battery voltage? Doesn't a low battery voltage mean that we're drawing too much current? I can tell this from the data logs, so while knowing exactly how much current each part draws is fine, it doesn't help me if we need those electric parts for the car to work. I suppose you're saying we should try to then find less current hungry parts?

For this year, we'll be having a play with PWM and SSR, to try to reduce load on the battery for longer (start fan off at 6V/5A, then ramp up to 12V/10A). We didn't run out of charge over a full enduro, so while the situation is not ideal, it works well enough for FSAE comp (but not a 24hr race). It just means the battery has to charged before every test day, and constantly charged at comp.

We also ran a Braille lithium battery for 2010, which claims to have useful voltage for 3 times longer the equivalent lead acid charge, so this may have contributed. They don't have much technical info on their website, so I don't believe that 100%, as I can't really tell how it works. While the battery is VERY light, its an expensive way to save weight (was a $400 management decision).

[RenM]

Originally posted by Rex Chan:

We also ran a Braille lithium battery for 2010, which claims to have useful voltage for 3 times longer the equivalent lead acid charge, so this may have contributed. They don't have much technical info on their website, so I don't believe that 100%, as I can't really tell how it works. While the battery is VERY light, its an expensive way to save weight (was a $400 management decision).

Perhaps you should have a look at LiFePos as they are much cheaper and can usually deliver more current. Plus they are also very light. The reason Lithium Batteries are more useful, is that they can deliver high current even though they are almost discharged where as the usable current for lead batteries significantly drops when they are not fully charged.
A Lithium Battery with 10% charge will still be able to Start your engine while a Lead Battery may need 50% or more (just guessed values for illustration)

[Jan_Dressler]

Originally posted by Kirby:

I'd be surprised if any teams do separate to this unless they run one of the pace designs with the integral pressure stage and adjustable regulator.

I think Rex was more commenting on (and I'd tend to agree with him) when designing your dry sump pan to bolt to the engine, you have to be rather accurate in any sort of connections that engage internal passages or pickups. PRV location, Internal Pressure Pump Inlet etc.

Especially tricky with the R6's as the internal pump pickup is inclined on a 15degree incline to the sump flange, so something that is able to engage that and line up on the dowels...is a challenge.

We tossed our engine upside down on a CMM and built a fairly accurate model of the internal pickups, which formed the basis for cast dry sump pan design.

Fun times..
Ah okay, maybe I misunderstood... Agreed, naturally it is necessary to get good measurements (CMM, as you mentioned) of all the critical points.

[Jan_Dressler]

Originally posted by Rex Chan:
With regards to the PRV, we had (2009/2010) the actual valve inside the pan, but the oil flowing back into the oil tank, near the bottom. A few years ago, they dumped the excess oil back into the dry sump, but the scavenge pumps couldn't keep up. So, OP, try to avoid that if you can.
We have the PRV flowing directly back to the (stock) pressure pump inlet in a separate "channel" in the dry sump oil pan (i.e. no additional oil line to the oil tank, and no dumping into the dry sump --> no more load on the scavenge pumps + hoses). I think that's the best solution unless you have a pressure pump with variable flow rate (no need for PRV). http://fsae.com/groupee_common/emoticons/icon_smile.gif

Originally posted by Rex Chan:
I agree that -16 kevlar braided hose is overkill for a breather, but we used it for a couple of reasons: we had some lying around, and it was part of the oil system (so we used known oil + temp resistant hose, because not all hoses can take 115C). The silliest thing was that we secured it via hoseclamps at the oil tank end.
Actually we used a (I think) -8 kevlar braided hose as an oil tank --> catch tank hose on our last cars because of the very same reason: We had some lying around http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

Originally posted by Rex Chan:
About the specs on the Dailey scavenge pump: I could be wrong on the specs. I just remembered that the 3-stage (actually only used 2) Pace pump was 6L/min/1000RPM, while the bigger 2-stage Pace was 10L/min/1000RPM, and that the Dailey was somewhere inbetween. What I can be sure of is that it was the one from this page, for pre-07 CBR600RR: http://www.daileyengineering.com/motorcycle.htm

I am quite sure that the Dailey pump you mention (we use the same, 2x 1.05 scavenge sections, with air separator) has a flow rate of 16l/1000rpm per section.

Originally posted by Rex Chan:
We run electric water pumps for the same reason most FSAE teams who run dry sumps run them: the stock water pump shaft is being used by the scavenge pump. Jan: by the way your question was asked, I'll assume you think we have too big a water pump. The electric pumps don't like backpressure, and their flowrate falls off dramatically. We found a UNSW/ADFA paper that measured the flow of the Suzuki GSX-R600 mechanical water pump, which showed flow rates up to 65L/min @ 9000RPM, which is close to our average RPM operating range for our CBR600s. This would mean an FSAE car running a MWP could flow around 65L/min. However, the EWP does not pump anywhere near that amount of water, due to the backpressure in our system. So they don't even match the stock pump; and are not close to being oversized. As a result, we run a much larger than stock radiator.

Admittedly, we have an aim engine temp of only 85C (due to the EWP controller we run), which is lower than the stock thermostat of fully open 90C (btw, we have removed the thermostat, otherwise the restriction would stop water flow completely).
We did run a setup with the EWP80 and a radiator which was also a bit on the larger side (but not that much) last year, with no cooling problems whatsoever. I have to mention that we also run a air-oil-radiator, not the stock water-oil heat exchanger, which according to our experience is a more effective solution.
As I mentioned, we have a much smaller water pump than the EWP80 this year, with a slightly increased water radiator size. This solution was found to be lighter than keeping the EWP80 and decreasing the radiator size.

Jan

[Moddin]

Actually we used a (I think) -8 kevlar braided hose as an oil tank --> catch tank hose on our last cars because of the very same reason: We had some lying around Big Grin

No Jan, it was -6 http://fsae.com/groupee_common/emoticons/icon_wink.gif. We used one of the (forbidden) fuel hoses made of teflon of our '09 car. I think the world's lightest and most expensive hose for the catch tank.


Btw: Hi @all, I'm Martin, as Jan member of High Speed Karlsruhe. Last year I developed our dry sump further and did the fuel stuff. This year I am the team leader engine & drivetrain.