How to make a watertight seal?

[LegoLord1880]

Recently I'e been doing some vacuum engine experiments. My engines always get more performance from using lubrication, but I'm wary of using olive oil, since it left a really gross residue on some of my pieces. Eventually I realized water would work just as well, but would need to be applied often because of how things it is. I came up with a splash lubrication system, which uses a sealed crankcase filled with water that is splashed up onto the cylinder walls when the piston goes down. It works great, but the water leaks out from between the bricks and constantly needs to be topped off. I've been using saran wrap between the layers of bricks, but I guess that's not enough. Does anyone have tips on how to seal the crankcase?

Also, I would appreciate it if anyone has a water pump design to use for the water cooling system I developed. It needs to be able to pump water through a pneumatic hose.

[aFrInaTi0n]

I am not sure if I am of help anyhow, but I just thought of some sprayable rubbers / there are some manufacturers offering products for cars.

Name of one product is:

Spoiler

Plastidip

My idea would be, it could be applied between the bricks and any standoffs can easily trimmed away - et voila you should have the gap between both parts filled with a hardened rubber on mostlikely the whole touching surface area between the two parts.

But I never tested this for myself and it may be a stupid idea.. unfortunately wisdom for stupidity of an idea can just be found out with trying the idea.. 

[LegoLord1880]

Hmmm. Seems like it could work, but I don't want to do anything that could permanently stick the pieces together. Maybe I'll try more plastic wrap, and between each layer. Previously I had used only a double layer on the baseplate in the bottom of the oil pan.

Spoiler

 

 

[engine1ear]

Quote

I came up with a splash lubrication system

This is a really cool idea to have an active lubrication system for a LEGO air powered engine. At one point in my LEGO engine building I wanted crankshafts to be larger in diameter than just a LEGO axle so I made some roller bearings that I shared more about here. But to be more true to real internal combustion engines, it would be so awesome to have the bearings be plain oil lubricated.

Quote

I just thought of some sprayable rubbers

At one point I bought something of the sort because I wanted to seal the inside of a leaky engine head that didn't seal air well enough for the engine to run, but picturing the mess, I couldn't bring myself to use it. It might be expired by now.

I'm excited to see a LEGO active lubrication system with an "oil pump"! A quick search in the Technic Builds and LEGO Engines Discord server showed the engine with the paristaltic oil pump made by user snagman in 2023 (not my creation).

Hope this gives you some ideas!

Edited December 22, 2025 by engine1ear

[LegoLord1880]

:D That's a great idea, I'll have to try it sometime! It does look like it'd draw quite a bit of power though. Since we're both lego engine innovators, I think Id better give an update on what i did and the new designs i have:

The engine I did this one is a "high performance" 4x4 svs engine. Keep in mind I didn't use ANY performance designs like lightened pistons, cranks, or flywheels. Just a standard svs that I got up to 4000 rpm. You can see the mostly watertight oil pan in the picture, and the fill port where I put water in for lubrication. I use water because it makes less mess than vegetable oils, plus it doesn't leave residues, and it's thin enough to work well at high speed. The engine has splash lubrication for the bottom end, where the crankshaft dips into the water and flings it onto the lower cylinder walls. The seran wrap gasket isn't perfectly watertight though, so I have to run the engine in a 9x13 glass pan borrowed from my mom.

My engine also has the water cooling system I devised, and here's where things get a little complicated. I use a pnuematic pump submerged in a glass of ice cold water to pump water through a small radiator I made. The radiator is pressed against the engine block, allowing the cold water inside to absorb minute amounts of heat from the engine. And here is the last of the 3 innovations on this engine: water injection. Water coming out of the radiator is injected into the engine's intake, which lubricates the top end. Then, the water either drips past the piston and slightly refills the oil pan, or gets sucked into the wet-dry vacuum on the power stroke.

All the water action happens intermittently (bc of the manual pump), and the engine speeds up like its on nitrous when I pump water into the intake. It all works pretty well! I would love to see this stuff get put to use in other, more realistic engines (I'm talking to you, engine1ear! 🙂). I'll get some pictures uploaded and put in a link in a minute.

Water Cooled Engine pics: https://drive.google.com/drive/folders/1U7hSHggFk5SBkjLve5Ibm4_0fHKgKr5v

[allanp]

I'm not sure how useful this comment is, but reading this topic reminded me of how air is itself used as a sort of lubricant in some applications. I'm thinking of grinding machine spindles that run on a cussion of air to prevent metal on metal contact. The precision required to use air as a lubricant in this manner is much higher than is possible in Lego. However, doesn't a hovercraft work the same way? With enough air flow and the proper design, the mating surfaces can be as imprecise as a rubber skirt on a grass field. I wonder if a flow of air could be allowed to pass from the high (atmospheric) pressure side of the piston through channels that exit through ports on the side walls of the piston itself, creating a cussion of air between the piston and the bore in which it slides.

It could be true that the small bypass air around the piston is already at atmospheric pressure and thus no air flow will come from the jets. However, when air is forced to speed up through a small opening (such as air that is forced to blow by the piston), its pressure actually drops. The smaller the opening, the greater the pressure drops (Bernoulli's principal), and so I imagine the piston gets sucked towards the side walls increasing friction. So (in my head anyway!) maybe allowing a small cussion of air to exist via ports on the sides of the piston would not only act as a lubricant but also help prevent a drop in pressure that can suck the piston against the side walls?

All this is purely speculation. I'll leave you the fun of testing it 

Edited January 7 by allanp

[LegoLord1880]

That's a good thought, and I can see where you're coming from, but lego engines really need liquid lubrication, just like real engines. Because of how leaky the seams between bricks are, we can't allow more leaks than necessary. The optimal engine does use air as a cushion to some extent though. When building a vacuum engine, you actually have to stretch the cylinder walls away from the piston a little so the engine can turn more freely. Then on the power stroke when the valve opens, the vacuum pulls the cylinder walls against the piston to make a seal, and THEN the piston is sucked up. When the valve closes, the cylinder walls relax, break the seal, and allow more free movement. We are dealing with around .0015 inches though, so the seal doesn't really "break", but there is a noticeable difference running an engine without cylinder stretching. This is partly why vacuum engines tend to not run as well on positive pressure.

Also, like in gas and Diesel engines, the lubricant can be used to make the seal. The lubricant can "fill in" pitting or rougher surfaces to create seals, on top of making everything run smoother.   

[engine1ear]

Wow you actually went and made a radiator prototype! This is awesome man. Back in the days when I spent hours living in CAD world (I used LEGO Digital Designer) where LEGO didn't have gaps that could leak fluids, I made a radiator design that I was really proud of using ladder pieces to simulate the cooling fins. It's biggest flaw, as mentioned, it would have never held water haha. I went and dug up a photo.

For LEGO radiators overall, it's too bad that ABS plastic is a thermal insulator (0.14-0.21 W/(m·K) thermal conductivity) as compared to say, aluminum (200+ W/(m·K) thermal conductivity) so we'd have a hard time really getting much heat transfer. It would be great to gain some data on the concrete benefits of cooling a LEGO vacuum engine. In general I think they run cool enough not to melt anything until you get to thousands of rpm, then particular places seem to show repeated failure (just from my observations of mostly other people's engine failures involving melted plastic over the years) in the crank-side of the connecting rod. I would call that the big end, but when your conrod is a Technic liftarm...

So your choice of SVS engine means your cooling system might actually be mitigating this sort of damage/failure - nice! Regardless, I think there's a bright future ahead for innovations like those mentioned in this thread. Speaking of which,

Quote

Since we're both lego engine innovators...

I really appreciate that so much! Definitely wear that badge with pride!

Thanks so much for the updates on your innovations. This was really really cool to see, and I look forward to whatever you come up with next, whatever that may be! Feel free to reach out anytime!

Quote

...creating a cussion of air between the piston and the bore in which it slides...

I've never thought about the contribution that the blow-by in a LEGO vacuum engine has on the whole system to reducing friction. Thank you for that interesting thought experiment!

Cheers guys!

[AdamFMX]

This may be unappropiate (sorry if it is spelled wrong) but what about to try 3D print some sort of seal and cylinder sleeves to prevent leaks from piston part of engine?

[LegoLord1880]

Engine1ear has actually done that on his "life sized Lego vacuum engine" video on yt. I think its a great idea to have sleeves, but tolerances would be a major issue in getting it right, especially with how inconsistent 3d prints can be. 

1 hour ago, engine1ear said:

Wow you actually went and made a radiator prototype!

 

This was actually my SECOND liquid cooling system design. The original has two radiators and works way better. With that system, plain air is pumped through one radiator submerged in a glass of ice and brine. Then the second radiator is pressed against the block. The water in the cup is what absorbs heat, the ice is to make the water colder (absorb MORE heat), and there is salt in the water to make the ice melt faster (to make the water colder, to absorb more heat. it all works together nicely). So basically, air goes through the primary radiator to be cooled, then through the secondary radiator to cool the block and fill the "water jacket" with cold air. Like I said, this cooling system works WAY better than the one in my new high performance engine( mostly because of the ice water + salt mix, and because it emits cooled air instead of drawing heat from inside the cylinder). I ran this by my dad who's worked in the HVAC industry for 25+ years, and he said it would work reasonably well as long as there is still solid ice in the glass. I'm fine with that, since these Lego engines probably don't produce much heat tbh. But it still is cool to say I invented the Water Cooled Lego engine!

So why did I I use an inferior cooling system? Because I also wanted to use water injection and only have 1 pneumatic pump lol.

[2GodBDGlory]

18 hours ago, LegoLord1880 said:

Engine1ear has actually done that on his "life sized Lego vacuum engine" video on yt. I think its a great idea to have sleeves, but tolerances would be a major issue in getting it right, especially with how inconsistent 3d prints can be. 

In my opinion, 3D prints are pretty self-consistent. If you print the same thing several times on the same machine, it should be the same size. However, there can be inconsistency between different printers, so any parts would have to be designed for your printer's tolerances. It does sound like a realistic option to me, though! TPU is a rubbery material that can be printed, and I believe it works well for seals. It's also not a grippy rubber at all, so maybe there wouldn't be too much resistance to motion, either?

[engine1ear]

I believe the 3D printed sleeve I have is ABS because it appears the piston and cylinder are wearing at about the same rate. I'm not certain because it was printed for me when I was in University on a high end Ultimaker printer. But I'm pretty convinced it doesn't take any more than a Bambu labs level printer to get a workable surface finish. Then there's the idea of honing one of these sleeves with a ball hone very similar to how a real engine block is honed. I bought one of these hones intending to hone an SLS 3D printed cylinder sleeve, but the print turned out bad because of some trouble with the printer at work that day and I never printed another. I'd still like to do all that, but honestly the lower effort, the better. Print and go would be ideal for others to replicate. And that was the case with the sleeve I'm using pictured here. Now to find out if a Bambu printer can make a sleeve at that level of quality. If you're doing this, makes sure to set the seam alignment to random.

TPU is also an interesting choice and I definitely want to try it. I learned from another engine builder's video this week that PLA is not a good choice because it doesn't wear similarly enough to ABS so one would be destroying the other over time, apparently. I was going to try for ABS with my printer to get that even wear.

One more idea I've had is to try like a 30-45 degree print orientation just to get the layer lines something other than perpendicular to the direction of motion. Thanks for hearing me out guys, maybe you have some different ideas.

Quote

I also wanted to use water injection and only have 1 pneumatic pump lol.

That's really cool. I like how you positioned the "injector" right above where the engine sucks in the air. Perhaps, if you have a 4000+ rpm engine you could point it at the crankpin and run it full tilt to failure with and without the water injection system on. And THEN, you could compare the results to your splash lubrication system! That would make for such a cool YT video demonstrating the benefits of water cooling on LVEs - one I would LOVE to see! 

This is the type of connecting rod failure I mentioned in the previous post. Again, not my photo but from username stlxr on the TBLE Discord. Have you experienced this?

As a side note I've never had such an interesting discussion about LVEs on EuroBricks before! Thanks LegoLord1880 for initiating, and thanks others for your replies!

[LegoLord1880]

That's INSANE! I've never seen crankpin failure like that before! One time I had a piston weld to a cylinder wall though, and I keep both on a shelf as a kind of "trophy". Btw I've been trying to get into Discord to use TBLE but the stupid account verification is messed up somehow :p. I like your idea with lubing the crankpin. I just thought of a (complicated) way of doing that with my cooling system off the one pump actually. I'd have the air coming off the secondary radiator go into a sealed jar filled with water, and run a separate line out of the jar that would carry water. There is a really good picture of this in Sariel's "Unofficial lego technic builder's guide" book. Then I would tee off the water line and send one hose to the water injector and point the other at the crankpin to lubricate it. 

Edited January 8 by LegoLord1880

[engine1ear]

Quote

I've been trying to get into Discord to use TBLE

Oh, hopefully you can overcome the technology gremlins. It's a great community. LEGO vacuum engine building has had a renaissance in the last five years and the builders are mostly teenagers, so for the general EuroBricks audience I'll say join TBLE or other LEGO engine Discord servers if you don't mind being in a room of middle/high schoolers. No shade at all, of course. In fact the LVE community could really use some support from AFOLs with lots of experience in things like advanced building techniques as well as life advice and all the other positive outcomes that come from intergenerational encouragement. I've seen really wholesome interaction on TBLE over the years with the couple of AFOLs that have been active. And there's a lot of raw talent.

Quote

a (complicated) way of doing that with my cooling system off the one pump

Let's see it! Build and test, and even if it doesn't yield an obvious performance improvement, it doesn't mean you haven't learned something along the way!

Quote

seal the inside of a leaky engine head

Going back to this. This week I discovered a LEGO engine innovator's video who came up with a really simple way to improve the seal of a LEGO engine head that is not messy at all. A piece of paper. It not only covers a bunch of gaps between bricks, but also gives the "combustion chamber" of the engine head a smoother surface for air to flow against. Certainly better than that surface littered with anti studs!

[Need For Bricks]

Hello guys, I'm need for bricks, I'm the guy from the video that engine1ear linked here, he recommended me this forum and I got quite excited, I loved the radiator idea, and... Would've you mind if I used it in my own engine?  You can see it here: https://drive.google.com/drive/folders/1VYexszGyFS8NAILu0BMTU-wZN-PYwbFG

I was inspired by engine1ear's realistic engines and gave it a shot

 

Also please ignore my profile picture in the drive link I've sent

 

Cheers! Kauã 

[LegoLord1880]

It looks like you already did lol. Go ahead! :)

I've found it works best if you have multiple "water jackets" around the cylinder for more cooling. Sometime I'll have to borrow my dad's thermal imaging camera to see how big a difference water cooling makes

[Need For Bricks]

Actually the water didn't work, but I'll try an air cooling system, like you did in your first cooling system, I will use a pneumatic pump that pumps air to an air tank submerged in water with ice, it cools the air and the air is pumped to the radiator, and to an "air injector" on the head

 

Hope it works, also any updates on your engine? Thanks 

[LegoLord1880]

By water jacket, i mean places to mount the radiator against the block. they dont actually hold water lol. I'm gonna try making that engine again as an I2.

oh, and you really do need a radiator submerged in water. An air tank will have WAY less surface area and will also lower the pressure in the air line

[Need For Bricks]

Oh I get it now

 

I'm exited! You should try a dry pump oil distribution system like mine, they work awesome and no leaks (except when injected on the piston, there's a leak on the chain side of the piston) 

 

Good luck 

[engine1ear]

You guys have me really wanting to try out a system that uses water, in my case I'm thinking for lubrication. I liked seeing that photo you shared on a YouTube community page, NFB. I've noticed that the big end of my connecting rod is wearing away at the crankpin. I knew this would happen since I didn't design that interface with any kind of lubricated or roller bearing, it's just dry. You can see the crankpin I'm talking about in this video. The problem that really makes me consider water lubrication is that as it wears out it creates plastic dust which I think is increasing the friction and slowing the engine down. I'll have to do more testing of this theory, but it might be worth some development along the lines of what you guys have been working on. Thanks for sharing all your ideas, y'all!

[Need For Bricks]

Hello engine1ear

 

Actually I'm using a mixture of detergent and water as lubrificant in my engine, it's thin and actually cleans the cilynder walls, I recommend it for sure!

 

Also while trying the radiator thing I realized my pump won't work, neither for water or air, but I couldn't find a good design on yt that I had the parts or that didn't modify parts, so do you have more pictures of that engine with peristaltic pump for oil? Or so you know any other efficient design? Thanks 

[engine1ear]

Hm, I agree with what LegoLord said earlier that a paristaltic pump probably uses up a lot of the engine's power to run and may even just prevent the engine from running altogether. For a first attempt, if it were me, I would settle for a drip-lube system, then let things progress from there.

[Need For Bricks]

@engine1ear I mean... It would be driven by an electric motor... Clearly my engine has no torque, and an engine driven water pump would make the engine just not to run

[engine1ear]

Oh OK haha. It would be very cool to have an engine driven oil pump! Maybe one day someone could achieve a net zero power gain because all the power gained from the lowered friction would be used up powering the pump haha!

[LegoLord1880]

Or making it run off intake vacuum like the windshield wipers on old cars. THAT would be cool!