product categories


CSF Triple-Pass Radiator: Airing It Out

CSF Triple-Pass Aluminum Radiators have been incredibly popular with our customers, and we’ve got our fair share under the hoods of a few of our personal street and track cars around the office. Due to the complexity of a triple-pass radiator and the nature of the internet rumor mill, we feel the need to “clear the air” and offer some helpful details about these high-performance units.

Triple-pass radiators are extremely effective but require a bit more diligence when it comes to bleeding the system; if not bled properly, issues may arise. As an example, we ran across a forum thread from an owner of a CSF-equipped BMW who had a problematic cooling issue where the temperature gauge was showing a higher-than-normal temperature (purchased elsewhere without the BimmerWorld technical experience to lean on), and he ended up removing a “bad” unit and replacing with stock. One of our BimmerWorld customers reported nearly identical issues a few weeks later, but based on the description, we knew an incomplete bleeding process would very likely have similar results to what was being described. But not wanting to take any risks, and seeing their was a prior unresolved and reported issue, we immediately had our customer’s radiator returned to CSF for a thorough investigation.

The summary of the examination performed by CSF was comprehensive… They inspected the entire system, running a wire scope through the tanks with no findings, and they even went so far as to chop off the tanks for further analysis. The meticulous examination by CSF showed there was no clogging anywhere in the system and all tubes were open with adequate flow.  Result?  As expected, there was air in the system from an incomplete bleed.

So, why are we sharing this information? It’s not purely to show how seriously we take potential issues and how great CSF is to work with (but we are certainly proud to sell a range of hand-picked parts from these types of suppliers that “get it”), but primarily to illustrate the learning experience – and pass it along! With the returned unit passing all tests, this simply proves that the CSF Triple-Pass Radiators require more effort in the bleeding process than standard radiatorsWe’ve even had the same issues here, despite knowing how complete the bleeding process must be.

The fact is, because of the 3-pass design’s longer flow distance and vertically segmented compartments, the bleeding on these particular radiators takes more time and diligence in order to get all the air out of the system. This part of the process is critical for the triple-pass radiator to be able to function as advertised – and it does! We’ve had countless numbers of customers with data showing a consistent 25-30% reduction in temperatures on track.


The longer flow distance of the CSF Triple-Pass Radiator increases the potential for trapped air that must be removed for proper operation


CSF Triple-Pass Radiator Bleeding Recommendations

1) Perform full radiator bleed with the car’s nose elevated
2) Backfill the system through the rear hose until it bleeds freely through the front
3) Get the engine up to full operating temperature, then park it overnight to completely cool before re-bleeding
4) Repeat as needed

While the bleeding process is crucial for all radiators, this level of thoroughness is an absolute requirement for the triple-pass, 2-row systems. As long as the system is fully bled correctly, the CSF Triple-Pass Radiator is an extremely capable and reliable cooling solution, particularly on track.  Additionally, while we often use a specialized vacuum bleeding system, and especially for difficult systems, it is important to note that use of this tool alone (without following the above process) will not guarantee a proper bleed.


Notes For Street Use

The cooling performance of the CSF Triple-Pass radiator makes this part an ideal choice for track cars, and it’s almost overkill (but perfectly acceptable) for high-performance street use.  For those driving in hot ambient temperatures, especially with air conditioning on, heat soak from the AC condenser can affect the last third (in the 3-pass configuration) of the radiator core where the fluid exits. The radiator is working correctly in these conditions, but because of the location of the temperature sensor, the triple-pass configuration is not going to produce a cooler outlet temperature versus a single pass unit – and that is not the intent of this radiator! Despite the factory gauge reading being elevated, temperature probe testing shows that temperatures are well within the range for safe operating (205°), even with the AC on at full blast.

E46M3 CSF Front

The CSF Triple-Pass Aluminum Radiator is designed to aid your car on track in the most demanding conditions, and has no negative effect versus common street parts in the street environment – and all in a cost-effective package.


Is a Strut Brace Worth It?

When it comes to car modifications, strut tower braces have been the topic of many a great debate on the Internet. Whether for street use, track use, or additional chassis reinforcement on an already stiff and caged race car, the question always comes up: Is a strut brace worth it?

Yes, you need a strut brace!
“Yes! You need a strut brace. Trust me! You’ll feel it working!”

There are those who will argue to the e-death that they can feel a noticeable improvement in chassis stiffness and suspension control with a strut tower brace installed, and there are those who refuse to even look at strut bars because they believe they’re nothing more than engine bay bling and eye candy. There’s plenty of conjecture, but it’s extremely hard to find any actual facts to back up claims from either side. Since we love to have actual data to make an informed decision or recommendation, we turned off the proverbial butt dyno and decided to measure the actual amount of strut tower deflection on a BMW. What can we say? We like to go the extra mile – and we like learning just as much as we like sharing knowledge.

No, strut braces don't do anything!
“No! A strut brace is a worthless piece of eye candy. I couldn’t feel anything!”


When in doubt, test!

Spec E46 testing
Our Spec E46 330i was the test subject

While plenty of people like to run strut tower braces on street cars, we wanted to get measurements on something that would see much more severe duty, so our test car was our BimmerWorld Spec E46 BMW 330i, and we used both Lime Rock Park and the full course at Virginia International Raceway (VIR). VIR gave us the most usable data since it has elevation change and some significant high-loading features that really put a chassis through a workout. Plus, it’s our home track, and we’re intimately familiar with the layout.


Testing was conducted without a strut bar in place to determine “the problem” that a strut bar would otherwise solve. Nick Large, BimmerWorld’s Product Development Engineer, configured the onboard AiM MXL2 data system to include linear potentiometers in several locations/configurations to measure distance changes between points. After recording on-track data, we filtered out some level of noise with an exponentially weighted moving average, or for those of us who aren’t engineers, that means data smoothing, which gives us a cleaner look at the data. You can see the results in the graph below.

BimmerWorld SE46 Strut Tower Deflection Data from VIR
(Click image to view the full-size graph)

The Delta Distance on the graph measures the change in distance between the tops of the strut towers, so a positive number means the towers are spreading apart while a negative number means the towers are getting closer together. Note that the measurements are in millimeters, or more specifically tenths of millimeters. As a rule of thumb, the distance increases during cornering because the load on the tire is twisting the top of the inside tower away from the vehicle center line, and the distance decreases under braking and over bumps because the towers twist simultaneously and in opposite directions. If you know the layout of VIR well enough, the annotations allow you to understand the trends and see how they correlate to certain sections on the track.

The data was gathered over many laps, with every lap showing incredibly consistent deflection numbers, and as you can see, the movement in this measurement location is minimal on this particular car. A strut bar, or any other structural member, provides strength almost exclusively in this tension/compression range, and most E30/E36/E46 BMWs exhibit movement purely laterally across the strut towers, as our measurements reflect. Further testing of additional locations showed similar minimal movement, although those locations wouldn’t typically be affected by an aftermarket strut bar regardless.


So, is a strut brace worth it?

That brings us back to the main question: Is a strut brace worth it? The scientific answer is, “It depends!” Based on the data from our specific Spec E46, this car didn’t need one on this particular test day.

But that isn’t the comprehensive answer, and from experience, we know there is a need in some cars. Why is that? BMW is iterative in their chassis design, and part of the added weight of each new chassis is the added stiffness. The older, lighter, and more flexible E30 definitely needs a strut brace, which we know from the high number of racers with collapsing strut towers that are actually pushing together with age; a strut tower brace solves this! And one hard concept to accept is that everything wears out over time, including the metal chassis of a car, so even newer models will likely head down this path eventually with enough mileage, age, and pounding on the street or track.

We’ve also seen other E46 cars bend strut tower braces, which was the catalyst for this experiment, so there is clearly something moving. But remember, standard strut braces help in tension and compression, not bending, so while there is movement, a strut brace really didn’t help much in the cars where we saw damage. And why was there movement? The example cars we saw had been hit previously in the front, weakening the spot welds or sheet metal. If our test car were older, in worse shape, or if it had any broken spot welds in the strut tower, then a strut brace would help.

People ask why we don’t recommend strut braces for all cars. Easy answer – not every car requires one, and we don’t believe in pushing products you don’t actually need. But there are certainly situations where BMWs benefit from having a brace.

E46 strut tower brace

So, what about installing a high-quality, single-piece strut brace like the Rogue Engineering Race Brace to simply add reinforcement to the strut towers – the prevent a problem before it happens approach? That could certainly be a benefit, as it may very well help to disperse the energy from potholes, bumps, and severe cornering loads (especially when running camber plates). Anything you can do to help distribute loads and avoid deformation of the strut towers is a plus, but there are less expensive ways to do that if you don’t need a strut brace – BMW’s factory strut tower reinforcement plates being a prime example. The minimal movement shown in our test graph proves to us that a strut brace would be most beneficial if you have an older car or tired chassis with noticeable flex and/or pre-existing issues. Then you may get the intended benefits of the brace with the added insurance against strut tower deformation.

Ultimately, we’d like to perform this same test after five years of racing with this car. While the movement in our graph is minimal, it may certainly be enough to have a fatigue effect over time. According to James Clay, who once had his finger stuck between the main hoop of a cage and the B-pillar for part of a lap by accident (ouch!), this would likely be no more fatigue than the entire metal chassis of the car would suffer from. So it becomes a question of the added weight from an extra part (weight is normally considered the enemy) versus the potential long-term benefits. That’s not as easy to answer.

What customers are saying...