I work with water pumps so I feel obligated to point out that the reason the impellers had trouble starting up is because the reservoir inside the pump housing isn't filled with water all the way and has air in it. Water pumps are supposed to be "primed" before starting them which means filling the entire reservoir with water. Otherwise the impeller is trying to push air out when it's meant to push water. So this explains the difference in results after you did the second round

Today I learned that yellow color filament will give me the best performance for 3dp impeller

The strait blade impellers are high volume low head designs, the curved blades are lower volume higher head designs. That is why pump performance is graphed as a curve and prefers to operate within a specific duty cycle between open flow and deadhead (blocked discharge). You can simulate feet of head pressure (psi×2.31=) by throttling the discharge with a valve and monitoring psi with a gauge. Motor load must be figured into your pump's performance as well. Great video and happy engineering!

This isn't a positive displacement pump so it has to be below the water line to function best. Also it would be good to measure the head of each design with just a tall pipe and marking how far it can push water up. Good video though

Yellow and grey impellers looks like they were design by glorius german engineering xD

Being no engineer, I learned that 1) Some designs need priming before use or they choke on air for ages (but perform well) and 2) some designs are FAR better at self priming than others (but don't perform as well). I'm interested in knowing why, but that would likely involve some slow motion footage and possibly an animation per impeller. Cool stuff!

I'm very glad that you did both high pressure low volume with the hose fill to 5L and high volume low pressure with pumping straight out. Very interesting the difference in results! Great video!

You would have totally different results if the pumo was set on the bottom of the reservoir, gravity would prime the pump quickly

Wow was the previous installment of this really uploaded a year ago? Wow how time flies…I remember watching it like it was yesterday! I’m glad you came back to run this experiment again!

The 1st test with the longer tube for resistance and the 2nd test which was more of an open flow gave me plenty of data to judge pressure vs flow. Thank you!

Keep in mind that pump efficiency depends on motor power too, you should measure motor current and voltage during each impeller test. An impeller that moves water a little slower but consumes much less power is more efficient. Keep up the good work!

Mechanical engineer here, different shapes of impeller can have different performance curves (perform differently at different levels of back pressure) ideally each propeller should be tested at three different levels of back pressure or restriction in your nozzle and then all those points should be plotted and a curve can be fitted the compared against each other so that the best impeller can be chosen based on application. This would explain the different performance in the two different tests as one test had a hose with a high restriction and one test had a nozzle with low restriction. Anyway great work!

The airfoil on the black impeller is travelling in reverse. That fix should see major improvements.

Great video! It would be interesting to see how each impeller operates under "pressure". As well as impeller design and its effects on cavitation.

I was just amazed by the purity of the experiment... and incomparable embodiment of details... and the attention to the sealing of the motor axis was especially amused ...

An idea to make better seals is to put "grooves" into the border of your prints: make a few layers of valleys and peaks. This increases the surface area of sealing. Nice video!

Not disclosing sponsor :angry: Also images of the impellers on the chart would be nice for reference

*Major Hardware* another KZfaqr has same kind of content but he does with PC fans... I really love your videos and make more designs and probably even make a page to submit viewer submitted impellers, print and test them!!!

My guess for those new objects is that they are for testing archimedes screw designs. :)

it would be cool if the pump motor current is also measured.

Nice video. Thanks for your efforts. I would also be interested in the power used by the pump with different impellers. Maybe next time.

Bro add valve on the intake and prime the engine with water removing all the air before actually measing the performance. What you did at the end was actually pretty good idea to sort the problem. The video is amazing. Keep it up!

Excellent video, well made and explained, thanks ... Please note, never use grease or any secondary form of substance on your gaskets, use gaskets only... You get leaks and will cause gaskets to perish in time, put a metal cross bar over the two top bolts to get extra pressure on that leaking top straight... Nice one... Respect to you... Thom in Scotland.

You gave us your best choice ! I hope a good next clip. Thank you so much!

Next time i suggest using a tub to catch the water from the impellers so you dont waste water and it is easyer to refill.

may many more years come full of creativity and innovation, keep up the good work eh?

I feel like the green was doing better initially before it siphoned air back during the output change from open to the hose. Same with Black I wonder what results would be if you allowed the pump to be fully primed without air. (You can see air bubbles traveling through the hose, showing a massive inefficiency) 👌

The weak point is at 3:26, that o-ring will be grinded to dust after a few hours, so the lifetime is very short. But very nice video. Look for sealings they use in dish washers and washingmachines

5:00 For the gray frame cover, try to reinforce it with a middle 90 degrees web to avoid bends also try to find a way to add one more screw to the top at the middle. good luck with the rest of the video

We use the blue design (albeit with 6 to 8 vanes) with our solder pumps for electronics manufacturing. I wonder how well these other shapes would do and how the weight of the liquid would affect the result.

Congratulations on getting your pump to prime.

Hello, interesting test, I recommend that you set the pump lower than the deepest point of the container with water so that the pump is always filled with water and there is no air in it. A design like the one used for turbochargers would be interesting, and take a slow-motion recording (some cell phones can do that) to test the cavitation effect.

Appreciate the effort here ! Some more suggestions in addition to all the experienced folks here 1) all the leaks need to fixed to have more accurate results 2) why is the inlet pipe at a slant in the bucket ? It should continue to start from the bottom, but it should be horizontal all the way till the impeller. Even this isn't ideal as the water weight/pressure will reduce as the tub gets emptied. So I'd think we need a larger/wider tub so that height drop changes very little. 3) what improvement can we get if the inlet to the impeller isn't perpendicular rather inline with the impeller thrust ?

This would be really interesting if you measured the pump curve of each impeller and compare

i love your content i always loved to work with water end pumps when i was a little child so appreciate it :D

As far as the seals go, this type of pump leaks a lot. We use them at work, they require expensive, delicate fiber gaskets with a large amount of clamping force and a separate multi stage seal for the motor shaft that uses centripetal force and a separate seal fluid to turn back any liquid that leaks out the motor shaft

I love these vids. keep up the great work.

Now have a vertical clear tube that goes straight up a few meters and see which pump can move water the highest. You may find that the fastest at moving water in these tests may not be the best one in that situation.

very good idea to do such testing, I was expecting there will be some result related to power consumed for each design. But still good idea.

I have an idea (which may already be a thing). Variable pitch/angle impeller, i bet like a regular Propeller, they have different effectiveness at different speeds and i wonder if changing the blade angle could improve its performance

for pressured water pumping, the blue one works well. For an open water pumping, the yellow one works the best.. So this test shows what is best for every kind of situation.

Well, this comparison video is stupid, the pumps are not well primed, there are leaks everywhere, and the battery will loose its juice the more you use it. I would prefer to use dc power supply that provides stable voltage and high current

I wonder if the battery voltage had some part to play in this. It's possible the battery was charged between runs or that the draw wasn't enough between runs to make a difference though I'm not exactly sure. It would be interesting to do a similar test but instead using a different power source like a psu or something.

I would recommend coloring the water if you do another video like this. Great video though.

I really enjoy your video, few suggestions: Place the pump bellow the plastic tank that way it will be prime and results will be more reliable. Perform at least three test with each design. Place more screws to avoid leaks Thanks it was entertaining

seems like these in an R/C boat would perform differently because being submerged outlet/inlet. i think the high flow full trash/ sediment design might work better than the high pressure low sediment water only . because duck weed or sand or whatever but then nozzles definitely change for highest top speed vs highest acceleration torque

Awesome video, I need to make one to help me drain the pool when it's time to pack it up. Also I had to pee twice from the water sound lol

The gray one looks like a windmill of friendship and acceptance

Interesting.. experiment can be done base on volume & pressure output.. in the 1st experiment (1kg test) the rubber tube variations ( height, curve) can cause diff variation result even with the same colour impeller..

Love the videos, just wanted to say that having all tests run simultaneously on-screen may help viewer retention/reduce fast forwards. Unfortunately that would make your videos shorter, and I also know next to nothing about running a successful yt channel. Thanks again for quality content and solid methods tho

I like it I think it’s a cool video a cool to learn how different designs effects the effectiveness also I have a suggestion if you do this again when you are starting the pump and the water begins to shoot out I think you should have a bucket or something there to catch the water that way it will be easier to refill the water box

Hey. I love your videos. Like other comments say, these pumps are not self priming. So need to be below water supply level.

Beautiful & informative 🎥🔴 video 👌 .. The only thing i would suggest is priming the water pumps & then time the gallons litters per minutes or sec.

Yeah.. But then there is the volute geometry.. The inlet and outlet position. as well as the cut water position with respect to the impeller.. You could also consider some 3D impellers. Or impellers that have angled blades. Sweet video!

I think which impeller is best depends on other aspects, like motor power used fluid, etc.

Thank you very much for your sharing, very nice demonstration :)

To provide leaking you have to make the holes in the transparent cover bigger then the screw diameter

Greetings, 1st I wanna say to keep up great work there and great content. I am like 30+ years experienced mechanic and technician and my profession is hydraulics and pneumatic systems and equipment. I do lots of other jobs but that one is, lets say, main profession :D It would be great to see differences between static and working pressures of each impeller type and how much is the load on the electric motor for each impeller as well for each state. Few suggestions though... Try to make an groove on the top spot of the pump casing (where is the leak) and put one peace of bridge on the case, and make new seal in one piece that is enclosed all around. Also, try to fill the pump 100% with water before starting the test. It is called Priming, most of new pumps have that feature automatically soon as you open the intake valve, but rest of them need to be primed, older pumps. Or, you can make yourself Automatic priming system if you want to keep the pump above water level. It is easy to make as well, a self-priming centrifugal pump has two phases of operation: priming mode and pumping mode. In its priming mode, the pump essentially acts as a liquid-ring pump. The rotating impeller generates a vacuum at the impeller's 'eye' which draws air into the pump from the suction line. One more thing, to be able to fully test the impeller shapes, you will need to print the impeller with an EYE in the center of impeller which is going inside the intake pipe, it is the most common shape that prevents bad water suction, directs the water flow in right direction inside the pump and impeller itself and have huge impact on pump performance as well. (in the video, just have a notice how water behave inside the pump case and how much unnecessary turbulence are made inside impeller area, those are because bad impeller and case design) Anyway, it is great to see the differences regardless the imperfections keep up great content

Try testing a more aggressive design similar to a car tubine impeller! :D but very similar to the best performing one to see if you can squeeze out any more performance out of it!

and you know what, you should make a o-ring track seal for the housing, think that would work way better. And yeah, as other has stated, better test to make the pump not needing to suck water up into the housing. Maybe print a container on top of the hosing with a valve on the outlet of the pumphouse

Plz try this!!!! You should really try Black, Grey, and yellow with a Tri-Blade design against the blue and red traditional impellers. The larger space between blades to catch water may make a difference with the curved impellers as they can catch and hold more water to push out during each rotation. But if the blades are too close it creates the opposite effect and can't achieve its maximum efficiency.

use rtv with your gasket and you will have no more leaks...keep in mind the cure times. also, maybe use flat washers to spread the compression

i know leaking our tradition, but solving the leek also good :D

If you do another test I would like to see the static pressure differences between each design. I'm curious to see if maybe that would be why the black is the common design when others seem to out perform it in other aspects.

Nice video! This was a flow pump performance (no pressure). You should also compare pressure using a long tall transparent hose and see which impeler produces the tallest water column. Cheers from Argentina.

LW-PLA is often better for water seals. It foams and fills gaps between layer lines that cause leaks. That may sound counter intuitive because some foams absorb water, but that's only open-cell foams. Closed-cell foams do not. It is more expensive but if you want a watertight seal it's really the right choice. Just keep in mind it gets a lot more stringing than regular PLA.

Pretty clear why the pump leaked. That gasket didn't cover the top portion where water was spewing. I probably would've used an oring type gasket and some grease or gasket maker to ensure a good seal. Another test would be to test mixture of best prime/speed. Without proper sealing they're all invalid IMO.

6:42 next time, make the water level higher than the pump itself, but the air is sucked 🙏😂👈

Great video!! But: You should test some positive arrange of the blades all the designs but the first (wich is 0) are negative. Also, waterpump has two big problems, first one is that the higher part of the tube is the waterpump, wich could make air enter to the system. So you should add a gooseneck at the end. Second problem is that the waterpump should be oriented horizontally to avoid some 9.8ms2 forces.. I will be waiting for the third one.

This is like "Major Hardware Fan Show down" but for water.

Great work sir!

The impeller is strictly calculated by mathematical formulas. And its shape is not arbitrary; it is verified based on calculations. This is done by students at any technical institute during course design and then defend their project before the engineering commission. My course work was called - Calculation of the main parameters and the impeller of a centrifugal pump. Approximately 3 pages of mathematical formulas.

Thank you, i thing an important test for the engineer 👍🏽

You should try doing some FEM/FEA before printing out 18391839183929 different impellers, there’s also the Euler’s model for tubo machinery

Hi, great job! Which nozzle you used for printing? Thanks

Be interested to see the same test but with a 1 metre lift I bet the black impeller does better with more pressure.

Both of these were volume tests. I want to see a pure pressure test. Pump the water into a vertical pipe and mark how high the water level gets with each impeller.

Very interesting. But wouldn't case design affect the performance. Lets say the casing had the outlet port radially from the center, instead of forcing the water through the curved path. Would the black impeller perform best again? Also would like to see the efficiency readings.

Best impeller is S find sandwiched between two plates intake plate has center hole of about 1 to 1-1/2 inches. I know this because I use to work on industrial pumps. And you need ceremic pump seals on the shaft and top of impeller plate to seal around the shaft to illiminate leaks around the shaft. As far as casing leaks you need to use silicone on both sides then sandwich it together with screws and silicone over screw heads.

This is like the fan showdown channel. Nice!

The motor should work on its rated power so you can select the best impeller by that along with higher water output.

if you use a high dura tpu for the impeller and reduce the tolerance between the blade and the wall so they just touch you will get more efficient output and less turbidity in the flow.

this could benefit from a small wall on the lid piece and a indent in the body and "clear" case like this >>>, nested. (edge ring, clear part, body). Especially the high pressure around the outlet. Should increase rigidity.

The snail shell is what really does the pumping

i wonder also the number of impeller blades will impact performance. i need to cfd a lot of this

Thanks you so much, so helpful!!

Thanks for the video.Excellent.

Does painting yellow also work?? (Btw, Good experiment. Show the impeller in a graph so that I don't need to skip back to see which impeller you are referring to. Also, you can use a bench supply so we can track ampere it will tell more about how hard is it working )

Hey, you could use another reservoir to catch and reuse the pumped water. Apart from that nice video 😁

Hey I really liked your video I have been looking for a video like this with 3D and pillars for a long time and I really like how you changed your test up at the end it made it a lot more consistent. Great work can't wait to see more from you

You should also try anti-clock curved blades...

did you test max pooling depth and how far it can push water up? nice video keep the good work

I think each impeller design has it's own specific functionality. Idk.. Maybe you need to add pressure to make it efficiently pump or something like that..

Kinda crazy that the pump leaks from the only place where there is no gasket or bolts... Very surprising! 🤣

i assume so, but can this be used as a vacuum cleaner as well? i'm having a hard time finding impeller design guidelines for vacuum cleaner impellers

to make the black work better you must change the shape of the clear plate some on the pump just check out the design of the pumps made with that impeller... also you can put a throw foot on the blades. also might want to check out amp pull of the motor for each type...

very helpful, thank you

Which impeller would create a deeper vacuum in the box? Which impeller would have sucked more water if the box were closed?

Try to make the sealing surfaces flat, try to eliminate the 90 degree lip near the neck of the pump. Also, try to keep the clamping forces as consistent as possible. Given the design, putting another bolt in the middle of the neck would be impractical, try making the sealing band near the neck stiffer between those two bolts. Also, you may want to contour those surfaces when printed so that they are flat when the bolts are tightened. Have you ever noticed how empty flat trailers are bowed? Same idea.

El video arroja datos importantes en la interaccion caudal y tipo de alabe, sin embargo el diseño aunque luce atractivo necesita la utilizacion de sellos mecanicos y caucho para evitar fugas. buen trabajo