Possibly a quick tip for acetone smoothing that I haven't seen in the 3d print community yet, it is common to dehaze car headlights using an acetone mix vaporised in a electric heated mug with a plastic nozzle on the top. I believe these can be bought from your typical chinese retailers pretty affordably. Using this method would allow you to 'spot' vapor smooth the parts ie, only the internals while leaving the outside and critical details undamaged. This could provide the best of both worlds. If you like this idea and choose to use it could you please send me a download link for the supersprint files? Cheers, Blair
you might try printing it from nylon using jlcpcb or pcbway 3d printing services, they use SLS its much smoother and much more durable than any FDM print.
Would really be interested to see the difference from the 4 port nozzle to a single port and then adjust nozzle sizes for best flow vs pressure. This is something that's adjustable with sleeve inserts on full size boats. If it doesn't increase performance much than definitely not worth redesign thing steering.
According to the tests I did, the 4-path nozzle greatly outperforms any regular single path nozzle I have ever designed. I found this during the development of the M-Jet 35: th-cam.com/play/PLiFskGz_svvQjxSCElF1c1PSJKaHq8vae.html&si=X7RAr3htwXjIYbh0 The size of the outlets is already tweaked to achieve the highest top speed. There were attempts to change their size on the M-Jet 35, changing the size in both directions led to worse performance. This wasn't actually tested on the Supersprint, but the Supersprint pump was designed based on a scaled down M-Jet 35, therefore no significant achievable improvement is expected.
Would you ever print a Berkeley jet pump for fun? They are the most common endurance jet boat, and drag race jetboat reference design pump. They are also, very, very old designs adapted from what farmers once called "trash pumps". Called such because the old mix flow basic pump was once used for high volume irrigation, and to this day, you can still buy Berkeley brand irrigation pumps. I would love to see what you think of one in comparison to your amazing design. It would be great to see the side by side performance on a printable scale. While hamilton/scott/kodiak jet hav made some amazing advancements for jetboats, still none compare for raw thrust of the old mixed flow Berkeley. Maybe your design is the next step. I do not own a 3d printer, and honestly would not know where to start.
Thanks for your kind words and your suggestion! From what I have seen, the Berkeley pumps use a mixed flow impeller with an outside ring. There are multiple challenges with this kind of design, when attempting to combine it with some of the design elements I have been using so far. Firstly, when using two stages, which so far seems to be essential for priming ( th-cam.com/video/ZDD8SjKUy_4/w-d-xo.htmlsi=7kXQWqWu1EI8J-5S ), the berkeley conical shaped mixed flow design ends up having a bit more complicated pathway through the pump, changing the diameter back and forth rapidly. The diameter would have to increase through the first impeller, then around the first stator it would have to be decreased again to enter the second impeller. Apart from this geometrical complication, it is also important to understand what differences performance-wise there are between axial and mixed flow impellers. The axial flow impellers operate efficiently under lower pressure difference and higher flow rates, whereas mixed flow impellers operate efficiently with higher pressures but lower flow rates. That also matches the difference in geometry - axial design clearly allows more flow. When using two axial impellers in the two stage configuration, as seen in the video, it is the pressure that gets, in the ideal case, doubled. Therefore positioning two axial impellers in series might have a bit similar effect as using a single mixed flow impeller, with the advantage of improved priming. In fact, what I am currently using for the newest M-Jet pumps is a slightly mixed flow shape, which can be seen around 1:48. I found this makes the pump slightly more efficient. But converting the shape to even more mixed flow as 1:51 only worsened performance. Another complication which rises up with the Berkeley impeller is the ring around the impeller. This rotating ring needs to be placed inside the pump, inside of a stationary wear ring. The gap between the impeller ring and the stationary wear ring creates an annulus-like geometry. To prevent backflow in this area (there are no impeller blades to keep pushing water forward), the gap should be either somehow sealed or machined precisely to make it fairly small. A small gap, on the other hand, means that tiny debris or sand could get into the gap and eventually make the pump completely stuck. This might not be a problem for precisely manufactured Berkeley with smooth surfaces, but it could potentially create a problem for 3D printed surfaces. With these complications mentioned, I need to say that only a real test would show which exact effect a berkeley style design at such small scale would bring. I would be interested how the berkeley impeller changes priming, efficiency and overall behavior of the pump. To conclude - the Berkeley pump design is something I am interested about and might attempt to test some of its elements in the future, with some of the issues to be somehow overcome.
Join the official M-Jet Facebook group! facebook.com/groups/623045356071128 Supersprint files available for purchase: www.cgtrader.com/3d-print-models/hobby-diy/other/m-jet-sprint-3d-printable-rc-jetsprint-boat-model Supersprint hardware kit by RCPrinter: rcprinter.com/products/m-jet-supersprint-v1-0-build-kits
Right now it is not my focus, but it would be worth testing how much gain there can be from optimizing this parameter on M-Jet 35. I will also consider which improvements people report with this update on the Supersprint when actually running the boat. Primarily thanks to being larger, even the stock non-smoothed version of M-Jet 35 is still slightly more efficient than the smoothed Supersprint V0.4 pump.
@@mjetdevelopment Thank you for the input. I don't run rc boats but I have a lot of interest in using electric jet pumps on kayaks. Your work is some of the best that I have seen and I'd love to see something scaled up for small electric pwcs.
may I ask if you have tried changing the bearings and if you did which were the results (ie ABEC 7 vs ABEC 9, different supplier, ceramic or stainless...)? Great video!
Thanks! I am using stainless bearings and I don't know their ABEC rating. I haven't tried different ones, but I am convinced that they don't have any noticeable impact on the pump performance. The pump is a high torque machine, the pump in the video requires about 0.4Nm to be spinned. Quick calculation on SKF website, and the main bearing which takes all load will only produce about 0.00045Nm of resistance, making this source of power loss negligible.
Hello! It seems like a great improvement in terms of consumption :D. I have all the parts of v0.3 printed, is everything compatible? Can I print just the pump and the propeller to go with this improvement? Or should I also print the hull of the boat? Greetings! And thanks again!
@mjetdevelopment What parts are actually different from the v0.3 to v0.4? I’m in the middle of printing, don’t know if I should stick with printing the rest of the v0.3 parts or the v0.4. Would the v0.4 impeller work with the v0.3 nozzle housing? PS, thanks for organizing and updating the files in v0.4!
Perfect, I haven’t printed either of those yet :) Thank you! Your quick and helpful responses to questions from the community is very much appreciated, and makes this process more enjoyable and less intimidating. Thanks again!
Bro, kg of trust per kw of power that’s will be standard measurements and more perfect it know inlet power and outgoing power ( and get know the best efficient set volts/amps for your setup)
Yeah, plotting thrust per kw would also be a good metric and information. The only disadvantage of that metric is, that this does not include the exit velocity, which is very important for top speed. You can have 20kg thrust but very low exit velocity, then top speed will be low. Therefore I am plotting the relative efficiency, which results from output power / input power, compared to the Sprint. The output power already includes the exit velocity, too.
@@mjetdevelopmentdo yo7 know how a speed meter device looks like on bottom of getski? U can easily do something same, and measure velocity.. or if u somehow can collect all outputted water in 3 seconds with 5kg thrust, u can calculate a velocity.. e=mc2/2.. c2=e*2/m.. c*c=5kg*9.8*2/m
@@elyaperestigli2774 No need for that, I am calculating the exit velocity from thrust and exit nozzle area. The equations are shown at 0:38 here: th-cam.com/video/un8rdTVPTKM/w-d-xo.htmlsi=QLlE8AJtOyFXXgPo
Wow thats a big motor. I need to note that scaling anything up comes with many complications - hardware sizes (bearings, screws), electronics mounts sizes, wall thickness, tolerances, overall its a difficult task. But if you eventually get that motor, you can aim for 60-70mm pump diameter with that kind of power.
Yes, you are right, I hid the video. After I released it, I had some more thought around it. Right now the M-Jet Development is my only side job while I am studying. I really like it and I would like to continue. In order to keep my position on the market and my income, I can no longer give away all of my knowledge and methods. Anyone who comes to design jet pumps after me can copy what I do without having to spend those 5 years to figure it out as I did. As you can see from my videos, I like to share stuff. However, by doing it, at this point I am just putting my job in danger. Therefore I need to find a balance between sharing everything and keeping some tricks for myself. I will continue to share the results of my work; probably no longer everything which led to them. Thanks for understanding.
@@mjetdevelopmentCompletely understand your reasoning. It’s too bad though, I’ve been learning so much from you and really enjoyed that last video. It’s been very interesting following your progress. Keep up the great work, we all appreciate your efforts 🙌🏽.
BTW Stepan, even if you no longer share the internals of your data logging experiment, I would appreciate if you could at least validate how far off are my calculations in the spreadsheet I left at the RC Groups forums 🙏
About 200% for that motor size. But I definitely don't recommend scaling it unless you can do CAD and work with the STEP file. Scaling it up brings many problems with hardware sizes and more.
is the file for the pump bidirectional? I think i would like to try building a catamaran with 2 Pumps but i would need the Servo and the steering to be reversed/flipped to the other side ^^
You can just mirror all parts of the pump in your slicer. Then spin them the opposite directions. But keep in mind that with the Supersprint, the pump housing is completely integrated into the rear hull. So if you wish to use the pump for a catamaran, you would need to heavily modify that rear hull file or create your own housing.
hey there, i got a question, i just got an mjet-30 printed out and wat woundering if the propellers could handle 36.5k rpm, probaly a bit of a long shot though, i printed them at 100% infill with no supports if thats help. thanks, - TI0S_94
@@dfg1ml You can use any standard material like ASA/PLA/PET. You need TPU for the cover lid. Take a look at the manual: drive.google.com/drive/folders/1f-epKNBlcqGcuJU5rGvYQleL7ayPnrRe?usp=sharing
was waiting for a second version of the pump to start printing, here we go now!
0.4 pump will be printed tomorrow! Even the "old" one was a blast!
Possibly a quick tip for acetone smoothing that I haven't seen in the 3d print community yet, it is common to dehaze car headlights using an acetone mix vaporised in a electric heated mug with a plastic nozzle on the top. I believe these can be bought from your typical chinese retailers pretty affordably. Using this method would allow you to 'spot' vapor smooth the parts ie, only the internals while leaving the outside and critical details undamaged. This could provide the best of both worlds. If you like this idea and choose to use it could you please send me a download link for the supersprint files? Cheers, Blair
Thanks for sharing your research, this is very interesting. Can't wait to build a supersprint.
Amazing improvements, well done! 👍👍
Fantastic video! Probably the chart that you should create is the grams of thrust per watt of input power over the RPM range.
Thanks! Yes, plotting thrust per watts and including the whole RPM range would also give good information.
Beautiful work
I LOVE IT!
THX soo soo much.
Just waiting on the new printer to work...
you might try printing it from nylon using jlcpcb or pcbway 3d printing services, they use SLS its much smoother and much more durable than any FDM print.
Yes, good point, using a smoother printing technology will improve the performance even further
Would really be interested to see the difference from the 4 port nozzle to a single port and then adjust nozzle sizes for best flow vs pressure. This is something that's adjustable with sleeve inserts on full size boats. If it doesn't increase performance much than definitely not worth redesign thing steering.
According to the tests I did, the 4-path nozzle greatly outperforms any regular single path nozzle I have ever designed. I found this during the development of the M-Jet 35:
th-cam.com/play/PLiFskGz_svvQjxSCElF1c1PSJKaHq8vae.html&si=X7RAr3htwXjIYbh0
The size of the outlets is already tweaked to achieve the highest top speed. There were attempts to change their size on the M-Jet 35, changing the size in both directions led to worse performance. This wasn't actually tested on the Supersprint, but the Supersprint pump was designed based on a scaled down M-Jet 35, therefore no significant achievable improvement is expected.
Very cool
Would you ever print a Berkeley jet pump for fun? They are the most common endurance jet boat, and drag race jetboat reference design pump. They are also, very, very old designs adapted from what farmers once called "trash pumps". Called such because the old mix flow basic pump was once used for high volume irrigation, and to this day, you can still buy Berkeley brand irrigation pumps. I would love to see what you think of one in comparison to your amazing design. It would be great to see the side by side performance on a printable scale. While hamilton/scott/kodiak jet hav made some amazing advancements for jetboats, still none compare for raw thrust of the old mixed flow Berkeley. Maybe your design is the next step. I do not own a 3d printer, and honestly would not know where to start.
Thanks for your kind words and your suggestion!
From what I have seen, the Berkeley pumps use a mixed flow impeller with an outside ring. There are multiple challenges with this kind of design, when attempting to combine it with some of the design elements I have been using so far.
Firstly, when using two stages, which so far seems to be essential for priming ( th-cam.com/video/ZDD8SjKUy_4/w-d-xo.htmlsi=7kXQWqWu1EI8J-5S ), the berkeley conical shaped mixed flow design ends up having a bit more complicated pathway through the pump, changing the diameter back and forth rapidly. The diameter would have to increase through the first impeller, then around the first stator it would have to be decreased again to enter the second impeller.
Apart from this geometrical complication, it is also important to understand what differences performance-wise there are between axial and mixed flow impellers. The axial flow impellers operate efficiently under lower pressure difference and higher flow rates, whereas mixed flow impellers operate efficiently with higher pressures but lower flow rates. That also matches the difference in geometry - axial design clearly allows more flow.
When using two axial impellers in the two stage configuration, as seen in the video, it is the pressure that gets, in the ideal case, doubled. Therefore positioning two axial impellers in series might have a bit similar effect as using a single mixed flow impeller, with the advantage of improved priming.
In fact, what I am currently using for the newest M-Jet pumps is a slightly mixed flow shape, which can be seen around 1:48. I found this makes the pump slightly more efficient. But converting the shape to even more mixed flow as 1:51 only worsened performance.
Another complication which rises up with the Berkeley impeller is the ring around the impeller. This rotating ring needs to be placed inside the pump, inside of a stationary wear ring. The gap between the impeller ring and the stationary wear ring creates an annulus-like geometry. To prevent backflow in this area (there are no impeller blades to keep pushing water forward), the gap should be either somehow sealed or machined precisely to make it fairly small. A small gap, on the other hand, means that tiny debris or sand could get into the gap and eventually make the pump completely stuck. This might not be a problem for precisely manufactured Berkeley with smooth surfaces, but it could potentially create a problem for 3D printed surfaces.
With these complications mentioned, I need to say that only a real test would show which exact effect a berkeley style design at such small scale would bring. I would be interested how the berkeley impeller changes priming, efficiency and overall behavior of the pump.
To conclude - the Berkeley pump design is something I am interested about and might attempt to test some of its elements in the future, with some of the issues to be somehow overcome.
Have you thought about printing the pump in ABS like Resin or some other engineering grade resin?
I haven't, I don't have a resin printer. Presumably, for efficiency a resin print would be even better.
Bad ass!
Great video!
Join the official M-Jet Facebook group!
facebook.com/groups/623045356071128
Supersprint files available for purchase:
www.cgtrader.com/3d-print-models/hobby-diy/other/m-jet-sprint-3d-printable-rc-jetsprint-boat-model
Supersprint hardware kit by RCPrinter:
rcprinter.com/products/m-jet-supersprint-v1-0-build-kits
Any chance of testing these improvements on the m-jet 35? I'd love to see the performance and efficiency gains on the larger more efficient pump base.
Right now it is not my focus, but it would be worth testing how much gain there can be from optimizing this parameter on M-Jet 35. I will also consider which improvements people report with this update on the Supersprint when actually running the boat.
Primarily thanks to being larger, even the stock non-smoothed version of M-Jet 35 is still slightly more efficient than the smoothed Supersprint V0.4 pump.
@@mjetdevelopment Thank you for the input. I don't run rc boats but I have a lot of interest in using electric jet pumps on kayaks. Your work is some of the best that I have seen and I'd love to see something scaled up for small electric pwcs.
may I ask if you have tried changing the bearings and if you did which were the results (ie ABEC 7 vs ABEC 9, different supplier, ceramic or stainless...)? Great video!
Thanks!
I am using stainless bearings and I don't know their ABEC rating. I haven't tried different ones, but I am convinced that they don't have any noticeable impact on the pump performance.
The pump is a high torque machine, the pump in the video requires about 0.4Nm to be spinned. Quick calculation on SKF website, and the main bearing which takes all load will only produce about 0.00045Nm of resistance, making this source of power loss negligible.
Try printing the pump horizontal for a better surface
That would require supports which would end up worsening the surface
Hello! It seems like a great improvement in terms of consumption :D. I have all the parts of v0.3 printed, is everything compatible? Can I print just the pump and the propeller to go with this improvement? Or should I also print the hull of the boat? Greetings! And thanks again!
Yes! Everything is compatible, you only need to print the new V0.4 impeller. The pump is not changed.
@mjetdevelopment What parts are actually different from the v0.3 to v0.4? I’m in the middle of printing, don’t know if I should stick with printing the rest of the v0.3 parts or the v0.4. Would the v0.4 impeller work with the v0.3 nozzle housing? PS, thanks for organizing and updating the files in v0.4!
Only the impeller and the Lid_base is changed compared to the V0.3. Everything is compatible.
Perfect, I haven’t printed either of those yet :) Thank you! Your quick and helpful responses to questions from the community is very much appreciated, and makes this process more enjoyable and less intimidating. Thanks again!
Bro, kg of trust per kw of power that’s will be standard measurements and more perfect it know inlet power and outgoing power ( and get know the best efficient set volts/amps for your setup)
Yeah, plotting thrust per kw would also be a good metric and information.
The only disadvantage of that metric is, that this does not include the exit velocity, which is very important for top speed. You can have 20kg thrust but very low exit velocity, then top speed will be low.
Therefore I am plotting the relative efficiency, which results from output power / input power, compared to the Sprint. The output power already includes the exit velocity, too.
@@mjetdevelopmentdo yo7 know how a speed meter device looks like on bottom of getski? U can easily do something same, and measure velocity.. or if u somehow can collect all outputted water in 3 seconds with 5kg thrust, u can calculate a velocity.. e=mc2/2.. c2=e*2/m.. c*c=5kg*9.8*2/m
@@elyaperestigli2774 No need for that, I am calculating the exit velocity from thrust and exit nozzle area. The equations are shown at 0:38 here:
th-cam.com/video/un8rdTVPTKM/w-d-xo.htmlsi=QLlE8AJtOyFXXgPo
Im so tempted to scale this up into something ridable... TP Power 100 ???
Wow thats a big motor.
I need to note that scaling anything up comes with many complications - hardware sizes (bearings, screws), electronics mounts sizes, wall thickness, tolerances, overall its a difficult task.
But if you eventually get that motor, you can aim for 60-70mm pump diameter with that kind of power.
BTW Stepan, if I may ask: what happened to the recent video about data logging, did you take it down? I was looking for it and couldn't find it.
Yes, you are right, I hid the video.
After I released it, I had some more thought around it. Right now the M-Jet Development is my only side job while I am studying. I really like it and I would like to continue. In order to keep my position on the market and my income, I can no longer give away all of my knowledge and methods. Anyone who comes to design jet pumps after me can copy what I do without having to spend those 5 years to figure it out as I did.
As you can see from my videos, I like to share stuff. However, by doing it, at this point I am just putting my job in danger. Therefore I need to find a balance between sharing everything and keeping some tricks for myself. I will continue to share the results of my work; probably no longer everything which led to them.
Thanks for understanding.
@@mjetdevelopment Fair enough, thank you for the info you have shared so far, and I truly wish you success. 👍
@@mjetdevelopmentCompletely understand your reasoning. It’s too bad though, I’ve been learning so much from you and really enjoyed that last video. It’s been very interesting following your progress. Keep up the great work, we all appreciate your efforts 🙌🏽.
BTW Stepan, even if you no longer share the internals of your data logging experiment, I would appreciate if you could at least validate how far off are my calculations in the spreadsheet I left at the RC Groups forums 🙏
@@geekmidget Yes I will take a look.
What size do I need for a Hobby 56112 780 kV surpass engine...or how many % must used
About 200% for that motor size.
But I definitely don't recommend scaling it unless you can do CAD and work with the STEP file. Scaling it up brings many problems with hardware sizes and more.
is the file for the pump bidirectional? I think i would like to try building a catamaran with 2 Pumps but i would need the Servo and the steering to be reversed/flipped to the other side ^^
You can just mirror all parts of the pump in your slicer. Then spin them the opposite directions.
But keep in mind that with the Supersprint, the pump housing is completely integrated into the rear hull. So if you wish to use the pump for a catamaran, you would need to heavily modify that rear hull file or create your own housing.
Where can I get those pumps from 3.2% set
This is the 3.2% champion:
www.cgtrader.com/3d-print-models/hobby-diy/other/m-jet-sprint-3d-printable-rc-jetsprint-boat-model
hey there, i got a question, i just got an mjet-30 printed out and wat woundering if the propellers could handle 36.5k rpm, probaly a bit of a long shot though, i printed them at 100% infill with no supports if thats help. thanks, - TI0S_94
Strength of the impellers is usually not a problem. Which material did you use, and which motor are you using on which voltage?
@@mjetdevelopment pla+ filament And a 3650kv spectrum motor on 4s
which filament is suitable for this
@@dfg1ml You can use any standard material like ASA/PLA/PET. You need TPU for the cover lid. Take a look at the manual:
drive.google.com/drive/folders/1f-epKNBlcqGcuJU5rGvYQleL7ayPnrRe?usp=sharing
Tip. Try resin printer.
Yes, good point, it would be the best method as long as the resin is strong enough
@@mjetdevelopment i use SUNLU ABS like resin. its very strong