[kern417]

Originally Posted by weehe126

Wow, that seems like a lot more coolant for something that actually look smaller. The 60% increase in fluid volume alone should help. Hopefully the core actually cools better too.

The frame is similar but it is slightly shorter. However, the core is over 2x as thick. that's where the extra capacity comes from.

[kilogram]

Originally Posted by kern417

Stock System = 3.9L/1.0Gal
With CSF Exchanger = 6.2L/1.6Gal

Stock Heat Exchanger = 6.6lbs
CSF Heat Exchanger = 10.6lbs

I also diluted the coolant, so now it's closer to ~35/65 coolant to water ratio. With fluid you'd be around a total of 9-10lbs increase in weight.

Any thoughts on how you might differentiate between the effects of 60% more coolant vs. the bigger heat exchanger? The added thermal capacity alone should make for a pretty noticeable improvement. If anything, the capacity effect should manifest more in terms of rate of temperature change of the system, while the exchanger should affect the equilibrium temp of the system (though it will affect the δT/t since it sheds heat faster). Did you notice the system ever hitting an equilibrium from IATs? We might not have the data for this in our log streams.

[kern417]

Originally Posted by kilogram

Any thoughts on how you might differentiate between the effects of 60% more coolant vs. the bigger heat exchanger? The added thermal capacity alone should make for a pretty noticeable improvement. If anything, the capacity effect should manifest more in terms of rate of temperature change of the system, while the exchanger should affect the equilibrium temp of the system (though it will affect the δT/t since it sheds heat faster). Did you notice the system ever hitting an equilibrium from IATs? We might not have the data for this in our log streams.

Can't have a bigger radiator without more capacity, so no way to tell. I also diluted my coolant ratio, so it's somewhere between 30%-40% coolant and the rest is distilled water. The highest I've seen my IATs spike is just under 140F, so I don't need protection to 250+ that 50/50 provides.

A larger heat exchanger in our case will do a lot (in theory). When you floor it, the pump turns on and a surge of cool coolant goes through the intercooler. A larger heat exchanger means that there will be a longer period of cool coolant going through.

Unfortunately the nature of water is that it also takes longer to cool down. The *goal* is that the larger radiator will help reduce this period in cases like mine where the IATs tend to hover at 20F-30F over ambient afer doing a few pulls.

Still no data since it has been raining non-stop since Friday (yesterday had a few flash flood warnings). But hopefully Friday/Saturday I can do some logging and post that up.

[kern417]

DIY added to the OP

[OzBMR]

Originally Posted by kern417

DIY added to the OP

Nice job

[kern417]

Alright, here's the data. I'll make a video later with full comparison/review, but I want to post the log here first to get some feedback.

Method: 3 back to back pulls from ~40-140, starting in 3rd at 2500rpm and running 4th out the redline. I truncated the data to be 3000rpm in 3rd to 6900rpm in 4th. I know it's not perfect but I think the general concept is solid.

Ambient/IAT Delta = Difference between ambient temps and IATs
Intercooler Delta = Difference between intake temps before and after the intercooler

https://datazap.me/u/kern4179/stock-...at?log=0&data=

My main takeaways
1. The Amb/IAT delta are similar, but the temps were 20F hotter for CSF pulls
2. The intercooler delta increased with the CSF radiator despite hotter ambient temps (made the intercooler more efficient)
3. The rise in IATs during a pull is reduced with CSF
4. No matter what, temps increase in back to back pulls

Let me know what else you want to see. I also have data showing how temps increase when parked for ~5 minutes, as well as how temps reduce once you get rolling. Again, it's not super scientific so I just have it saved for now since it might not be useful. From what i saw, CSF cooled down IATs more quickly when you got going but didn't make a big difference in heat soak when parked. So overall a good upgrade for roll/road racing, but not sure how much it'll improve for drag racing.

[OzBMR]

Originally Posted by kern417

Alright, here's the data. I'll make a video later with full comparison/review, but I want to post the log here first to get some feedback.

Method: 3 back to back pulls from ~40-140, starting in 3rd at 2500rpm and running 4th out the redline. I truncated the data to be 3000rpm in 3rd to 6900rpm in 4th. I know it's not perfect but I think the general concept is solid.

Ambient/IAT Delta = Difference between ambient temps and IATs
Intercooler Delta = Difference between intake temps before and after the intercooler

https://datazap.me/u/kern4179/stock-...at?log=0&data=

My main takeaways
1. The Amb/IAT delta are similar, but the temps were 20F hotter for CSF pulls
2. The intercooler delta increased with the CSF radiator despite hotter ambient temps (made the intercooler more efficient)
3. The rise in IATs during a pull is reduced with CSF
3. No matter what, temps increase in back to back pulls

Let me know what else you want to see. I also have data showing how temps increase when parked for ~5 minutes, as well as how temps reduce once you get rolling. Again, it's not super scientific so I just have it saved for now since it might not be useful. From what i saw, CSF cooled down IATs more quickly when you got going but didn't make a big difference in heat soak when parked. So overall a good upgrade for roll/road racing, but not sure how much it'll improve for drag racing.

Thanks for testing and posting the data

Interesting results, looks like an average of about 11% improvement over stock when looking at the IAT delta over the 3 back-to-back runs.

IAT's still increase quite a bit in back-to-back runs, but as you say, less than stock would have with the same ambient temp.

In summary, do you think it's worth upgrading the intercooler heat exchanger? Would you do it again if you had a stock B58 car to start modifying?

[kern417]

Originally Posted by OzBMR

Thanks for testing and posting the data

Interesting results, looks like an average of about 11% improvement over stock when looking at the IAT delta over the 3 back-to-back runs.

IAT's still increase quite a bit in back-to-back runs, but as you say, less than stock would have with the same ambient temp.

In summary, do you think it's worth upgrading the intercooler heat exchanger? Would you do it again if you had a stock B58 car to start modifying?

I'm keeping it since the gains are there. I think it's beneficial if you're running over 18psi on stock turbo, especially in the summer. I don't think we really know what a "safe" IAT is for our engines, but seeing 110F+ on a 70F day was concerning to me. It very well could be within the DME's capacity to compensate for it, but I feel better knowing that I have a more efficient system.

I was also trying to force a worst case scenario with less than 30s between pulls. Every time I hit 140, hit the brakes, downshifted, and did another pull. This would be more similar to road racing where you're always either on the gas or brakes. For a few random pulls when normally driving or messing around with friends, the main benefit is that IATs don't increase as much during a pull. Personally, I'd probably switch to a stage 1/2 OTS map to keep temps down on a track day and keep the heat exchanger as additional cooling support.

I also think this proves that a heat exchanger can only do so much. The relatively small intercooler is working hard, reducing temps by over 200F at times. A larger intercooler will be a better upgrade to reduce IATs when you get past a certain point, although it'll likely be 4x the price of a heat exchanger. Time will tell when offerings are made available.

This upgrade will mostly benefit people that do roll racing and road racing. If you're drag racing, nothing beats putting ice on the manifold between runs. It may show a reduction in temps when going off the line because you need airflow for it to work, but it'll heat soak while you're in the pits.

Other questions I have/other's have asked
1. Would polished perform better than the black one I got?
2. How much of the gains are attributed to reducing the coolant ratio?
3. Will the gains be greater if I compared stock v CSF in hotter temps?

[Glemm1970]

[OzBMR]

Originally Posted by kern417

I'm keeping it since the gains are there. I think it's beneficial if you're running over 18psi on stock turbo, especially in the summer. I don't think we really know what a "safe" IAT is for our engines, but seeing 110F+ on a 70F day was concerning to me. It very well could be within the DME's capacity to compensate for it, but I feel better knowing that I have a more efficient system.

I was also trying to force a worst case scenario with less than 30s between pulls. Every time I hit 140, hit the brakes, downshifted, and did another pull. This would be more similar to road racing where you're always either on the gas or brakes. For a few random pulls when normally driving or messing around with friends, the main benefit is that IATs don't increase as much during a pull. Personally, I'd probably switch to a stage 1/2 OTS map to keep temps down on a track day and keep the heat exchanger as additional cooling support.

I also think this proves that a heat exchanger can only do so much. The relatively small intercooler is working hard, reducing temps by over 200F at times. A larger intercooler will be a better upgrade to reduce IATs when you get past a certain point, although it'll likely be 4x the price of a heat exchanger. Time will tell when offerings are made available.

This upgrade will mostly benefit people that do roll racing and road racing. If you're drag racing, nothing beats putting ice on the manifold between runs. It may show a reduction in temps when going off the line because you need airflow for it to work, but it'll heat soak while you're in the pits.

Other questions I have/other's have asked
1. Would polished perform better than the black one I got?
2. How much of the gains are attributed to reducing the coolant ratio?
3. Will the gains be greater if I compared stock v CSF in hotter temps?

What about the addition of a water wetter like Purple Ice or Redline Water Wetter to reduce surface tension and improve heat transfer?

This should assist the heat transfer at the radiator and also at the water to air intercooler.

[kern417]

Originally Posted by Glemm1970

what's funny is on supramkv twisted commented that it didn't help them, but they think it's because they are past the intercoolers effective range.

https://www.supramkv.com/threads/csf...e-2#post-82061

Originally Posted by OzBMR

What about the addition of a water wetter like Purple Ice or Redline Water Wetter to reduce surface tension and improve heat transfer?

This should assist the heat transfer at the radiator and also at the water to air intercooler.

I lost confidence in going that route since some people commented that the coolant has lubrication properties as well for the water pump. I may try it just for summer but not sold on it as a viable option yet.

[Solid67]

Originally Posted by kern417

what's funny is on supramkv twisted commented that it didn't help them, but they think it's because they are past the intercoolers effective range.

https://www.supramkv.com/threads/csf...e-2#post-82061



I lost confidence in going that route since some people commented that the coolant has lubrication properties as well for the water pump. I may try it just for summer but not sold on it as a viable option yet.

I saw Justin's comment about being beyond the capacity of the stock intercooler. It doesn't seem surprising considering he's running an EFR 9280 which is good for 1200hp.

When are we finally going to see aftermarket IC? Plenty of options for the S55 and yes I know that for the S55 it's just upgrading the A2W whereas for the B58 its Intake Manifold + A2W, but still you would think there would have been one or two options by now.

[kern417]

Originally Posted by Solid67

I saw Justin's comment about being beyond the capacity of the stock intercooler. It doesn't seem surprising considering he's running an EFR 9280 which is good for 1200hp.

When are we finally going to see aftermarket IC? Plenty of options for the S55 and yes I know that for the S55 it's just upgrading the A2W whereas for the B58 its Intake Manifold + A2W, but still you would think there would have been one or two options by now.

Yeah I think his last dyno was just over 800whp. Honestly, I think a stock turbo at 500whp would need it more than a pure turbo at 500whp. It really depends on how hard you're pushing your setup, because a bigger turbo will be more efficient and generate less heat at the same power level.

Regarding the intercooler, There needs to be a market for it. It would be awesome if we could just swap a core like the S55, but a likely $2k+ manifold will be difficult to sell.

[yokata]

Originally Posted by kern417

Yeah I think his last dyno was just over 800whp. Honestly, I think a stock turbo at 500whp would need it more than a pure turbo at 500whp. It really depends on how hard you're pushing your setup, because a bigger turbo will be more efficient and generate less heat at the same power level.

Regarding the intercooler, There needs to be a market for it. It would be awesome if we could just swap a core like the S55, but a likely $2k+ manifold will be difficult to sell.

If you are crazy enough price doesnt matter I will buy it instantly

[kilogram]

Originally Posted by kern417

Other questions I have/other's have asked
2. How much of the gains are attributed to reducing the coolant ratio?
?

35/65 has about 12% higher thermal conductivity than 50/50, and about 7% higher specific heat. It'll freeze at about 0°F though.
(https://www.engineeringtoolbox.com/t...ds-d_1260.html), https://www.engineeringtoolbox.com/e...col-d_146.html.

Might be worth running this mix as low as you can get away with based on minimum temps where you live. Water has about 2.3x the thermal conductivity as EG (and about 2x the specific heat). How hard is it to flush/drain the system? Might be a worthwhile addition to the track day prep list to swap out for water + wetter.


I think you're right that the IC is the bottleneck. If your thermal circuit looks like this:

Charge air -/\/\/\/- Intercooler -/\/\/\/- Coolant -/\/\/\/- Radiator -/\/\/\/- Ambient air

Flux is a function of δT across the boundaries, so there are only a few knobs to turn to increase flux across R1:
- Keep the coolant temp lower (greater δT)
- Increase the conductivity of the coolant interface
- Increase the convection coefficient at the intercooler/coolant interface

Convection is a function of boundary layer thickness. Do we know what the flow rates on that pump are? If the flow is laminar you could theoretically get some gains by increasing the flow rate until it's turbulent/fully developed flow inside the intercooler core.

Do we have any info on the intercooler core? Pictures? Measurements?

[kern417]

Originally Posted by kilogram

35/65 has about 12% higher thermal conductivity than 50/50, and about 7% higher specific heat. It'll freeze at about 0°F though.
(https://www.engineeringtoolbox.com/t...ds-d_1260.html), https://www.engineeringtoolbox.com/e...col-d_146.html.

Might be worth running this mix as low as you can get away with based on minimum temps where you live. Water has about 2.3x the thermal conductivity as EG (and about 2x the specific heat). How hard is it to flush/drain the system? Might be a worthwhile addition to the track day prep list to swap out for water + wetter.


I think you're right that the IC is the bottleneck. If your thermal circuit looks like this:

Charge air -/\/\/\/- Intercooler -/\/\/\/- Coolant -/\/\/\/- Radiator -/\/\/\/- Ambient air

Flux is a function of δT across the boundaries, so there are only a few knobs to turn to increase flux across R1:
- Keep the coolant temp lower (greater δT)
- Increase the conductivity of the coolant interface
- Increase the convection coefficient at the intercooler/coolant interface

Convection is a function of boundary layer thickness. Do we know what the flow rates on that pump are? If the flow is laminar you could theoretically get some gains by increasing the flow rate until it's turbulent/fully developed flow inside the intercooler core.

Do we have any info on the intercooler core? Pictures? Measurements?

I don't drive the car when it's below freezing, but my garage isn't heated so it's very likely to get very cold. I don't think it's very hard to flush the system, and considering I wasn't trying to lose more coolant than I needed to, I still lost half of the capacity. If I really tried to drain as much as possible, I could probably at least get 80% out of it. Definitely something to consider.

I don't know the flow rate of the pump, but I do know Paul set it to max duty cycle when demand calls for it. I don't think it will help either, though. You either need a bigger reservoir, or a bigger intercooler to further optimize the system.

There are a few pictures from SC_B5X teardown. It's removable, but I'm not sure how you would install a larger/more efficient one in the same cavity.

https://f30.bimmerpost.com/forums/sh....php?t=1699735

Originally Posted by SC_B5X

Well, I think I might have been the first person to blow up the stock B58 Intake Manifold. We were doing some testing and I overboosted to 37psi (Pure 800) and I believe it backfired into the manifold and busted it.

View post on imgur.com

[kilogram]

Originally Posted by kern417

I don't drive the car when it's below freezing, but my garage isn't heated so it's very likely to get very cold. I don't think it's very hard to flush the system, and considering I wasn't trying to lose more coolant than I needed to, I still lost half of the capacity. If I really tried to drain as much as possible, I could probably at least get 80% out of it. Definitely something to consider.

I don't know the flow rate of the pump, but I do know Paul set it to max duty cycle when demand calls for it. I don't think it will help either, though. You either need a bigger reservoir, or a bigger intercooler to further optimize the system.

I was poking through the B58 ROM editor last night and it looks like the control routine for the pump PWM signal is based on ambient temp vs IAT minus ambient delta. In stock form at least, it's always playing catch-up... with an ambient temp of 20°C, it doesn't start pumping until there's at least a 20° delta between IAT and ambient. Sounds like your tuner has modded that table (this is what MHD does with their "performance" cooling modes too). There's a lot of interesting stuff in the boost tables

There are a few pictures from SC_B5X teardown. It's removable, but I'm not sure how you would install a larger/more efficient one in the same cavity.

https://f30.bimmerpost.com/forums/sh....php?t=1699735

That's exactly what I was looking for, thanks! It looks like you could easily package a set of Laminova cores in the same space. They state an estimate of 2.5mm of core per HP, rough guess it looks like you could easily get 1200HP worth of core in that space, maybe more.

If @[SC_B5X] could get measurements of that core, I could throw together some rough laminova end caps in Solidworks to fit that space. It looks like it also runs a serial flow path through the stock core (Hard to tell for sure from the pics). Changing that to parallel or mixed flow would let the whole core work more effectively.

How does the stock mani seal? Is it bonded or bolted? Looks like it's bonded. Might be worth talking to 3M.. they have some urethane-based adhesives that might be able to handle re-sealing..

[kern417]

Originally Posted by kilogram

I was poking through the B58 ROM editor last night and it looks like the control routine for the pump PWM signal is based on ambient temp vs IAT minus ambient delta. In stock form at least, it's always playing catch-up... with an ambient temp of 20°C, it doesn't start pumping until there's at least a 20° delta between IAT and ambient. Sounds like your tuner has modded that table (this is what MHD does with their "performance" cooling modes too). There's a lot of interesting stuff in the boost tables




That's exactly what I was looking for, thanks! It looks like you could easily package a set of Laminova cores in the same space. They state an estimate of 2.5mm of core per HP, rough guess it looks like you could easily get 1200HP worth of core in that space, maybe more.

If @[SC_B5X] could get measurements of that core, I could throw together some rough laminova end caps in Solidworks to fit that space. It looks like it also runs a serial flow path through the stock core (Hard to tell for sure from the pics). Changing that to parallel or mixed flow would let the whole core work more effectively.

How does the stock mani seal? Is it bonded or bolted? Looks like it's bonded. Might be worth talking to 3M.. they have some urethane-based adhesives that might be able to handle re-sealing..

Yep, that's exactly what it does. That's why temps can drop breifly when you floor it, because the pump suddenly turns on and you get a surge of cool coolant flowing through the radiator. The only downside is once it's heatsoaked, you're stuck for a while.

The manifold is plastic welded. I'll have to look into the end caps, haven't heard of it. But the inlet/outlet are on one side and come out of the bottom. I don't think you can have it exit on the opposite side due to the accessories.

[kern417]

This is also probably a long shot, but I'm looking into rerouting the coolant path to use one of my aux radiators for the engine and integrate it into the intercooler circuit instead. Seems like it would be a fairly easy retrofit but I'm probably not considering something.

My car only came with one of the aux radiators, so I could either add a second one or just pull from the one I have now. More coming soon.

[kilogram]

Bah, I underestimated the size of the laminova stuff. Playing with it in Solidworks looks like half of what I though it was. You *might* be able to squeeze 12 245mm cores in there (~1100hp), but it would be tight and likely less efficient than their 2.5mm/HP suggests.

The aux radiator sounds like a solid plan. I think it might be reasonable to consider using the Burger water injection tank in the trunk as an auxiliary reservoir too (where you could potentially put ice), but you'd likely need a second pump to deal with the flow losses from the long runs of tubing.

[kern417]

Originally Posted by kilogram

Bah, I underestimated the size of the laminova stuff. Playing with it in Solidworks looks like half of what I though it was. You *might* be able to squeeze 12 245mm cores in there (~1100hp), but it would be tight and likely less efficient than their 2.5mm/HP suggests.

The aux radiator sounds like a solid plan. I think it might be reasonable to consider using the Burger water injection tank in the trunk as an auxiliary reservoir too (where you could potentially put ice), but you'd likely need a second pump to deal with the flow losses from the long runs of tubing.

Yep, I also thought of that. More capacity, and keep it out of the hot engine bay. A few things I'm struggling with:

- We have a return line for intercooler overflow that runs to the stock tank. So at a minimum we'd need to keep the stock tank.
- At a minimum I'd want to lower the position of the pump so it's below the height of the trunk floor. Otherwise I'd need to find a new additional pump that works.
- Concerned about finding lines and fittings that are large enough to support the needed flow.
- There are questions about running it inline vs on a switch, but a switch would probably be easiest.

Some may have seen Peter's setup where he built a custom crash bar with an integrated tank. He said that + the wagner radiator was a significant improvement.

[kilogram]

Originally Posted by kern417

Some may have seen Peter's setup where he built a custom crash bar with an integrated tank. He said that + the wagner radiator was a significant improvement.

Ok, now THAT's pretty f'ing clever and awesome...

Plumbing would definitely not be trivial. I think the best way would be to run the rear tank pressurized and have a stronger pump (N55 coolant pump? EMS/stewart warner Ford GT intercooler pump?) push coolant through it (easier to push through a restriction than to pull through it). But that would likely affect whether you could open it to add ice, since it would probably overflow with the stock res sitting higher than it.

The other thought I had was secondary core built into a custom charge pipe in front of the throttle body as a secondary cooler. You wouldn't have enough space for it to be the primary cooler, but it could take enough of the edge off the temps for the primary cooler to work effectively at higher HP levels. Looks like you could cram a Bell AW450025120 core in that space. Something like this: