FL5 Overheated on Track

[Gansan]

Yes. That's why people connect their secondary radiators in series with that branch. Even that mishimoto one taps in that path.

Tube diameters are a good indicator of expected flow rate for each branch. But tubes themselves aren't really a restriction compared to the heat exchanges on the same lines (or lack of!). For instance, we believe the coolant line inside the turbo to be pretty much restriction free. So it should present a high coolant flow even though the tubing diameter isn't large. And this coolant is going back into the engine hotter than when it left. A friend plugged this line and saw immediate improvements (I can't recall numbers). Was it because more water was redirected to the radiator? Or because that heat was being transferred to the oil instead?

I don’t think you can discount the effect of tube diameter. Tube diameter is really significant given it’s not like the engine has that powerful of a water pump. Given that all the flow comes from the one pump, the pipe diameter will be sized to put the desired proportion of flow to the right place, just like oil passages.

Of course you’re right that a heat exchanger will add to the restriction but one of the design goals of a heat exchanger is that it will be more efficient if it has less flow restriction. I suppose it would be informative to perform a test where you could put pressure sensors upstream and downstream of a heat exchanger to measure the pressure drop and do the same thing for a tube the same length as what the engine contains to compare them.

Actually instead of sensors you could just use some clear tubing to make a primitive manometer to measure this accurately.

Sponsored

 

[Vito.FL5]

There's another branch that goes to the expansion tank before returning to the engine. In the FL5, this branch leaves from the radiator hot side endtank. It's another path that bypasses the cooling circuit.



For more info:
https://www.civicx.com/forum/threads/alphadigitals-fk8-secondary-radiator-install.83468/post-1288712

Yes, but this is the one we cant get rid of. It’s importat to bleed the air from the system.

 

[Vito.FL5]

I don’t think you can discount the effect of tube diameter. Tube diameter is really significant given it’s not like the engine has that powerful of a water pump. Given that all the flow comes from the one pump, the pipe diameter will be sized to put the desired proportion of flow to the right place, just like oil passages.

Of course you’re right that a heat exchanger will add to the restriction but one of the design goals of a heat exchanger is that it will be more efficient if it has less flow restriction. I suppose it would be informative to perform a test where you could put pressure sensors upstream and downstream of a heat exchanger to measure the pressure drop and do the same thing for a tube the same length as what the engine contains to compare them.

Actually instead of sensors you could just use some clear tubing to make a primitive manometer to measure this accurately.

Well, yes, most of the flow go trough the radiator but the amount that circulate the parallel systems is not negletable. Anyone who turned on the heater at the track will know what I mean… it makes a HUGE difference on temperature, in my case its the only way to do 2 full laps.

 

[spectre186]

Have we found a way to get actual instead of computed oil temperatures? Is there a large variance between the 2?

 

[blueroadster]

I scanned through this thread but did not read every post. That being said, modern cars run coolant through parallel systems in order to get everything hot as quickly as possible. In the case of emissions, a common manufacturer preference is to run a hotter thermostat since a hotter running block burns more particulates before exiting the exhaust thereby reducing emissions overall but also increasing wear. The FL5 is cool in that respect as it is not running a hot thermostat.

The FL5 on the other hand like many newer cars uses systems piggybacking off of the coolant lines as a means to get components up to operational temperature. After a cold start for example, a parallel system is good to get the transmission up to operational temperature. Where parallel systems tend to later cause issues is once everything has warmed up and certain elements of the system experience more heat/stress over others. Ideal operational temperature points for oil, coolant and transmission fluids are not the same. Using a piggyback off of the coolant lines is not the best option in that particular case. That leads to one wanting to divorce the parallel system which is good but should be done so in a careful manner so as not to disturb design and sensor function. I would hope that anyone going down such a path takes that into consideration.

For a daily driver, the car should be good to go as is. For a track or upgraded car, one really needs to focus on heat dissipation and the cooling system itself first. Hondata or other logging tools can help pinpoint potential issues along the upgrade path. Going full steam ahead with changing many things at once will leave support behind as you have charted your own and different path using the stock system. If you have gotten to the point of wanting to move past that, then you are going to have to contend with big dollar items such as a MoTeC ECU (awesome but requires many sensor calibrations beyond the average driver's interest level), cryogenic treatment, dog gears, etc. depending on your goals.

That was a lot of yapping. I guess my question goes back to the OP's original intention of converting their stock FL5 into a track car. What concerns or limitations on the stock platform were targeted to improve incrementally to observe changes?

 

[blueroadster]

Have we found a way to get actual instead of computed oil temperatures? Is there a large variance between the 2?

Why do you "need" to see actual oil temperatures?

Simple solution would be to install a separate sensor wherever an oil temp reading is needed for datalogging and fed into another independent system for analysis. Oil temp only tells you one data point though depending on where the sensor itself is installed. Oil pressure throughout the system also needs to be considered.

The K20 engine is a marvel from an engineering standpoint. Ambitions and goals for its use require an understanding of design and limitations on what you want to do with it. For example, my previous 2006 Civic Si had a K20Z3 engine from the factory. To meet my HP goals from a reliability standpoint, I had to send the block out to ERL for it to be sleeved and headwork done by another well respected company. End result from a fluid dynamics analysis was that additional investment was required because the K20Z3 stock oil passages were too restrictive and needed another dedicated oil feed line to the head to make everything happy.

What are your goals and intended usage for the car?

 

[blueroadster]

Have we found a way to get actual instead of computed oil temperatures? Is there a large variance between the 2?

An alternative consideration would be...How would the existing electronics and their algorithms handle receiving real verses computed inputs from other dependent components? It all depends on the sender and receiver.

 

[Gansan]

Have we found a way to get actual instead of computed oil temperatures? Is there a large variance between the 2?

Computed temps doesn't mean it's an estimate. Even a sensor needs to have a computed temperature if you think about it.

A sensor uses a temperature-dependent physical property, and that property is measured and translated into a temperature. For example, you can use a thermistor, which changes resistance depending on temperature. You put a voltage across it, measure the current, run that value through a calculation, and you have temperature.

Oil itself can serve exactly the same purpose as a thermistor. The temperature-dependent physical property is viscosity, which is measurable by pressure at different rpms. Given engine rpm and pressure, you can reliably calculate the temperature of the oil, as long as the viscosity you are using is what the ECU is calibrated for. If you change the viscosity of the oil, the calculation will be off, so that's one reason to get a sensor, so that you can change oil types. But for stock oil, the "computed" temp should be reasonably accurate.

 

[yeaitsahonda]

Have we found a way to get actual instead of computed oil temperatures? Is there a large variance between the 2?

I installed an Oil Temp Gauge in my car temporarily for a track event this weekend. Will update with how close the factory readout is compared to the gauge. My car is stock with the only variable being that I have 5-30 oil in it instead of 0-20.

 

[PointByPatrol]

I've got a track day coming up in 4 weeks, and I'll be out there bone stock (with the exception of brake fluid and pads).

I'll run the heat in the car and remove the tray under the hood vent, and we'll see what happens!

 

[spectre186]

I've got a track day coming up in 4 weeks, and I'll be out there bone stock (with the exception of brake fluid and pads).

I'll run the heat in the car and remove the tray under the hood vent, and we'll see what happens!

you’re the man! I gotta make sure I support you, do you have a website for merch?

 

[Gansan]

I installed an Oil Temp Gauge in my car temporarily for a track event this weekend. Will update with how close the factory readout is compared to the gauge. My car is stock with the only variable being that I have 5-30 oil in it instead of 0-20.

Awesome! But if there's a significant difference then we should consider that could be the different viscosity at play until you test it again with other weights to see how the difference varies.

 

[PointByPatrol]

you’re the man! I gotta make sure I support you, do you have a website for merch?

I've just got my Black Flag Edition T-shirts so far. I have a website, but would need to add several more skews before I open it to the public. For now I just use PayPal

 

[yeaitsahonda]

Awesome! But if there's a significant difference then we should consider that could be the different viscosity at play until you test it again with other weights to see how the difference varies.

Awesome! But if there's a significant difference then we should consider that could be the different viscosity at play until you test it again with other weights to see how the difference varies.

Unfortunately I don't plan on getting that deep into it. I just want to know how accurate the OEM read out is so I can make a decision on an oil cooler. After the track weekend I'll be removing the gauge.

 

[keller]

I installed an Oil Temp Gauge in my car temporarily for a track event this weekend. Will update with how close the factory readout is compared to the gauge. My car is stock with the only variable being that I have 5-30 oil in it instead of 0-20.

Where did you install the sender? Honda's model estimates oil temp at the sump.

A few people did this comparison for the fk8, me included. I have a p3 gauge and an aem sender at the drain bolt. I had to linearize the calibration curve, but it's supposed to be pretty accurate for the range 80-130 C - less than 0.2% deviation. Well, logr is surprisingly accurate enough. I even considered removing the sender and gauge after 2 trackdays, but will keep them because I'm installing oil coolers. During road driving, it deviates less than 3 C. On track, I noticed Honda's model has a too short time constant. Oil takes a lot longer to get hot and then cool down. After this transient deviation right after starting a hot lap, it's at most 6 C wrong. The FL5 model could be even better.

Btw, at least in the fk8 the calculation doesn't use oil pressure. I suppose it's the same with the fl5, Vito could confirm it.

Sponsored