Do angled adapter fittings really reduce flow?
Probably one of the most quoted pieces of advice on building a liquid cooling loop is “don’t use many angled adapters, they reduce your flow rates”. So, is it true? Is it just a myth or is it a reality? You probably presume it is true, angled fittings will reduce your flow, but still, how much? A lot of unconfirmed information is circling in the liquid cooling community and most of it is based on “he said, she said”, with no certain scientific proof.
Luckily, that’s why we are here to try and explain, and of course, prove or debunk as many of these advices. This subject is fairly old, you can find forum threads on it dating back to 2010, and the issue is still being debated often. In the old days, yes it was true, angled adapter fittings were very restrictive, but nowadays, for example, we have the EK-AF Angled 90° fitting which is a high flow adapter.
But still, fluid hitting an obstacle and being forced to turn a sharp 90° should impact the motion of fluid and flow rates, right? Well, in this article we are going to provide you with an answer to the question “Do angled adapter fittings really reduce flow?“ At the same time, we are going to provide information about actual flow rate readouts to show you how much the angled adapters really impact your flow rates.
Our little test was conducted with an EK-D5 PWM G2 Motor (12V DC PWM Pump Motor) mounted on a pump reservoir combo. The flow measurement was done through an Arduino unit and a simple “no name” flow meter. By knowing the max flow rate of a D5 pump we can get some approximate readouts on the flow drop as we add angled fittings to the test loop. Although this test is not laboratory-precise, it’s a good approximation of what could you expect from overusing angled adapter fittings.
First, we attached only one angled fitting and connected all of the essential components using 15,9/9,5 EK-Tube ZMT tubing. The Arduino and the software were calibrated to give a readout of 1470 liters per hour at 100% PWM duty cycle of the D5 pump, since the theoretical flow of the EK-D5 PWM G2 pump is rated at 1500L/H. At 100% PWM cycle the pump was doing 4750 RPM, and if you want to know more about PWM and how does it work, just check out our blog post about it. Sticking to the subject, here are the first measurements of the flow drop as we reduced the pump speed based on 75%, 50%, and 25% PWM duty cycle.
The chart above shows us approximately how much flow rate drop can you expect just by reducing the pump speed since the loop contains only one angled adapter and a few regular fittings. Now, let’s add a couple of fittings and see what happens then.
By adding two angled adapter fittings to the loop we can clearly notice an average drop in the flow rate of 10% throughout the four ranges of pump speed settings. Let’s keep adding fittings and do some measuring again…
The decrease in flow rate continues and we can record an overall drop of 7% compared with previous measurements. We will move up from five angled adapter fittings to seven.
Using six or seven angled fittings in one loop is considered to be out of the ordinary. This usually implicates a big loop with multiple radiators and water blocks. With seven angled adapter fittings in the test loop, we recorded an average drop of less than 5% of the flow in comparison with the previous setup. We added two more angled adapters for our final fitting test.
This many angled adapters in one cooling loop would be a rare sight, but still, we needed to do this for the sake of science. Again, the results show an average of 5% drop in flow rate if we compare the results with previous measurements. In total, if we compare the flow rate results of one angled adapter fitting with the use of nine fittings, the average drop in flow rate is around 24%. Nine angled fittings are a bit much and a 24% drop in flow rate is not that big of a deal. But how does this compare to the flow rate reduction of some components, like a slim 240mm radiator, or a CPU water block?
Slim radiators are well known for being a bit more restrictive. This is consequent of the narrow flow channels through which the liquid travels to the radiator. Therefore, it is simple to draw a conclusion – a slim EK-CoolStream SE 240 radiator is more restrictive than those from PE or XE series. The average drop in flow rate on all four tested pump speeds is about 22% in comparison with the test loop with only one angled adapter fitting.
High-performance water blocks are always noticeably restrictive in comparison with radiators or other components because the liquid is forced through a narrow gap onto the cold plate. Since Supremacy EVO is a high-performance water block, the cooling liquid accelerates through jet plate’s nozzle and turbulently continues its path through numerous very thin channels that provide an extreme heat dissipating surface area. This is why the Supremacy EVO can ensure high-performance cooling even when the flow rate of the liquid in the loop is relatively low.
Before we draw a final conclusion, we would again like to attract your attention to the fact that this test is just an approximate representation of flow rate reduction that you can expect in various situations. The measuring tools that were used are cheap and low grade. Why did we use low-grade measuring tools instead of high-tech laboratory equipment? For reasons of science, so that users could replicate and confirm similar results at home.
The never ending debate on angled adapter fittings can now go to rest. Do angled fittings reduce flow rates? Yes. Will overuse of angled adapter fittings completely kill your flow rates? No. So the next time someone yells at you and tells you that you need to ditch as many angled adapter fittings as you can, just tell them to “keep calm and use the angled adapter fittings”. Compared with all the water blocks and radiators, a few extra angled adapter fittings will go unnoticed in any cooling loop regarding the performance. On the other hand, a few angled adapter fittings can make the tubing process much easier and improve the aesthetics of the build.