14 replies to this topic

[Project_1275_GT]

Water Injection System [Home made version]

 

I have got all the original Metro Turbo parts that most people discard as they are not really needed. Turbo ECU that controls only the boost pressure modulation, this senses the pressure of the system and uses a solenoid operated valve to bleed off some pressure. This is set to trigger around 4 PSI [also 4500 RPM]. The waste gate actuator is also set to operate at around 4PSI. So the original system allows the PCR valve to bleed off pressure keeping the wastegate closed up to the max 7.5 PSI [This is oscillated at 12 Hz setting]. Result is a slower more controlled boost. Knowing how this system works I intend to use the same ECU control but rather than the output from it direct to PCR valve it will operate a relay. The relay will then operate 1 - the bleed valve for pressure. 2 - The opening of the water valve. This should result in the engine running without water cooling up to 4 PSI, Then the bleed valve and water valve will open. 

 

I will setup a test rig for this just to test water flow through the nozzle and pressure required to get good atomization, I intend to place the water nozzle in the inlet to the Turbo as some benefits have been noted to lower the Turbo temperature. There is some concerns regarding water droplets damaging the compressor wheel. But no confirmed detail. So i will test it. I intend to install two very basic temperature sensors - 1. at the inlet area to the turbo. 2 - in the inlet manifold. This will give me some info on actual temps before and after. 

 

The reservoir for the water will be an old mini expansion tank, the pressure for the system i hope to use from the boost pressure. This is my main challenge to get a atomized spray of water at low pressure. If more pressure is required I will add a pump. Really hope i can do it without. The solenoid operated valve from the ECU will prevent syphon of water. The expansion tank is only small but i hope enough. It should only be used over 4500 RPM and 4 PSI boost. 

 

I will also install a sensor to check water level in the expansion tank. If the level runs low the relay operating both PCR valve and the water feed valve will be disabled. This will result in the boost being limited to 4 PSI. Hopefully this will prevent problems with running out of water and running high boost. 

 

If all of the above works I hope to be able to run slightly higher boost than normally possible on a higher compression engine. 

 

BASIC LAYOUT OF THE SYSTEM

 

 

 

This is not showing water level sensor. I have yet to work out the best sensor type. It will simply involve another relay preventing the live feed from the ECU to the valve control relay.

Edited by Project_1275_GT, 01 February 2021 - 07:37 AM.

[Turbo Phil]

You won’t get anywhere near the required water atomisation by simply using boost pressure to inject the water.
Injecting it pre Turbo especially will require extremely fine water droplets.
I’d certainly look at a 150+psi pump and a proper nozzle if you’re going this route.
Adding a small intercooler and associated pipework made no noticeable difference to lag when I first turboed mine 22years ago. Though keeping the compression ratio up will aid spool a bit your main issue is that you’re fitting a biggish turbo to a small displacement engine.

Phil.

[mini13]

yep, pressure is your freind here.

[Project_1275_GT]

I will make a test setup soon and start testing various nozzle types with different pressures to see what spray patterns i can achieve. I am not looking to inject a huge amount of water. Just enough to make a difference. But this is where a high pressure system would be most beneficial. To achieve any result may require more water. More water will almost certainly require more pressure. My only concern then is the reservoir size. If i can achieve a usable system with the idea above it would be good. If it looks like i will require a larger tank and vast amounts of pressure i may just buy a proper kit. No point in re designing what's already available. 

[Project_1275_GT]

Small update on this - 

 

I already had a window washer motor / pump ready for the rebuild of the car, I ordered a fuel injection nozzle with a 0.2mm jet and mackled together some fittings - Here is the result of the water jet. 

 

 Water jet.jpg   79.34K   1 downloads

 

The jet is very fine and as its 0.2 nozzle should not consume a large amount of water. I did video this but cant upload so converted to a picture showing the spray pattern. 

[Steve220]

Remember you'll be fighting the boost pressure in the piping. So if that's 10psi of water and you're running 10psi of boost... you won't get any water.

[Project_1275_GT]

Remember you'll be fighting the boost pressure in the piping. So if that's 10psi of water and you're running 10psi of boost... you won't get any water.

 

Im planning on placing this directly before the turbo compressor housing, that should result in air being pulled by the turbo and assisting with the pressure. I will also use a solenoid valve close to the jet to prevent syphon of water. 

 

Thanks

 

Rob

[Turbo Phil]

If you’re injecting pre compressor you need to ensure the water is as finely atomised as possible, it needs to be like a fine mist.
You don’t want to be sending large droplets at the compressor wheel.
There’s some cheap 80-100psi pumps on eBay.

Phil.

[Project_1275_GT]

If you’re injecting pre compressor you need to ensure the water is as finely atomised as possible, it needs to be like a fine mist.
You don’t want to be sending large droplets at the compressor wheel.
There’s some cheap 80-100psi pumps on eBay.

Phil.

 

I will take a look on eBay and see what's available. The washer pump idea was a stolen from Visards book. I checked the average pump pressure for window washer pumps and its around 38 PSI witch is respectable for such a small pump, but I am unsure if it can sustain constant running. But it has the advantage of being small and compact. I have also found a pressure valve specific for water injection that works form 2 - 10 PSI. The original plan was to use the Metro Turbo ECU. I will most likely test with the ECU first but with the option to install a valve intended for this job if necessary. Its hard to say how long the pump would run if setup on the ECU. Using the correct pressure sensing valve i suspect it would run much more and then require a better pump, also using more water. Its a bit of a balance of how much will i need it to run. 

 

The water being sprayed direct before the Turbo is always a risk. I can find reference to it possibly causing damage but nothing confirmed. I feel this would give the best atomisation of water to chambers and mix if introduced before the compressor wheel. This will also make more sense now why i am rebuilding my own Turbos. They may get destroyed. 

 

Im going to post three ideas below that are very different to the normal thinking. Some already trialled but not being produced and some my own adaptions of existing or old ideas. 

[Project_1275_GT]

Air intake to compressor housing 

 

So i spent some time researching air flow, sharp bends and hose types, both can cause turbulence affecting air flow to compressor resulting in slow spool. Knowing the hose connection position on a T3 on the engine its a tight fit. It will need a sharp bend. I also do not want my filter in the engine bay due to air flow and ambient temp, I am copying ERA turbo and putting the filter under the arch with extended hose to the front low down. This should keep incoming temps as low as possible. So back to the connection i am making, anyone ever seen K&N torque ram? Vizard tested this with great results for flow. But its a box that flows air without a bend through 90 deg.. Perfect solution for the turbo inlet. So i designed one - 

 

 K&N Torque ram 2.jpg   11.58K   3 downloads    7 trim.png   12.51K   4 downloads

 

Now the torque ram is no longer available, not designed for a turbo inlet. I have not so much stolen the idea but changed a basic principal to suit what i need. The inlet i have designed also has the water cooling nozzle housed in it, a temp sensor to check temps before compressor housing. This will also contain 1 more test piece that I will list next  

[Project_1275_GT]

Water injection direct before compressor wheel 

 

I have read plenty of info on this but seems no one has done it, or they just haven't detailed it that i can find. Spraying water before the compressor wheel may cause damage. But this will almost certainly be best for atomisation. So how to Prevent damage.. I started to research air flow through a static blade into a rotating blade. This is very interesting. I will not bore you with detail and simply explain its been tested and the results are good. 

 

 Pre swirl.png   121.06K   1 downloads

 

This is best described as a pre swirl. This will be placed directly before the compressor wheel. This is fixed and does not rotate. Water injection will hit this before the compressor wheel. Maybe this will better protect the wheel. To show the advantage of this the following pictures show flow through a turbo with and without this in place.

 

 111.png   172.34K   3 downloads

 

Above is with pre swirl

 

Below is without

 

 2222.png   160.17K   1 downloads

 

I have read lots on this and the trials and tests on turbo improved air flow and reduced vibration, load. Sounds to good really to not at least try it. This is incorporated into the air inlet 

Edited by Project_1275_GT, 03 February 2021 - 07:06 AM.

[Project_1275_GT]

Exhaust manifold and turbo turbine inlet AR 

 

I must admit this area has confused me a little. The AR of a T3 is only 0.25 - loads of people recommend various other turbo units but they are all starting around 0.36

 

This confuses me. If the T3 is too big why is the AR smaller that that of better suited turbos?? I am no expert in this area. But Ive taken a look and think maybe just poor, old design of the snail is one issue. But doing a little research into turbos I see twin scroll are highly rated. So please feel free to call me a fool. 

 

Looking at the exhaust exit of the metro turbo manifold its 3 vertical slots, it then enters a huge opening of the turbo turbine inlet. This means a controlled single exhaust stream goes from manifold exit diameter to HUGE turbo exhaust inlet, result, pressure drop. This is my idea - 

 

 Picture18.png   66.36K   2 downloads

 

Hope this makes sense. It will bolt between the exhaust manifold and the turbo exhaust inlet. The slots match the exhaust side, The extended sections will not contact the housing and will be shaped to folow the profile of the inlet. 2mm min clearance all round. This will extend the gas flow right into the AR housing. This should improve flow speed and therefore improve spool time. 

 

Who knows 

[Project_1275_GT]

All above ideas are test ideas, the inlet seems to be a suitable solution to a difficult hose connection and may improve the flow by reducing turbulence, it also has the ram effect due to exit from the housing into the turbo compressor wheel. Plus it keeps spray nozzle, temp sensor and pre swirl all in one place. bolts on with two M8 fixings to the original compressor housing. If i had a lathe and miller it would be easy. I have a hacksaw and a battery drill. 

 

The pre swirl seems like it could give me two benefits, one to protect the compressor wheel from direct water spray and two to improve air flow into the compressor wheel. As this will be easily installed or removed i think its worth looking  at. I have added a link below to research on this. It is a little full on with calculations but it does indicate this is worth a look. 

 

http://isromac-isime....finalpaper.pdf

 

The last exhaust manifold idea is my own and logically must improve air speed of gasses into the smaller end of turbo snail. If the gasses move faster the turbo spins quicker. Faster spool. 

 

All of the above can be fitted easily and removed easily, there seems to be good logic in the ideas but when its done a RR session will be needed to prove any gain or loss. The only mod that will be a pain is the exhaust divider. Wont just be able to remove on a RR session. 

 

Yes I am on very strong medication. 

 

 

 

 

[Steve220]

Water injection direct before compressor wheel 

 

I have read plenty of info on this but seems no one has done it, or they just haven't detailed it that i can find. Spraying water before the compressor wheel may cause damage. But this will almost certainly be best for atomisation. So how to Prevent damage.. I started to research air flow through a static blade into a rotating blade. This is very interesting. I will not bore you with detail and simply explain its been tested and the results are good. 

 

Pre swirl.png

 

This is best described as a pre swirl. This will be placed directly before the compressor wheel. This is fixed and does not rotate. Water injection will hit this before the compressor wheel. Maybe this will better protect the wheel. To show the advantage of this the following pictures show flow through a turbo with and without this in place.

 

111.png

 

Above is with pre swirl

 

Below is without

 

2222.png

 

I have read lots on this and the trials and tests on turbo improved air flow and reduced vibration, load. Sounds to good really to not at least try it. This is incorporated into the air inlet 

 

A few threads to read, if they help.

http://www.methanol-...formation_id=54

 

https://themotorhood...-meth-injection

 

https://www.pistonhe...sp?h=0&t=680867

 

https://www.engineba... Injection.html

[Project_1275_GT]

 

Water injection direct before compressor wheel 

 

I have read plenty of info on this but seems no one has done it, or they just haven't detailed it that i can find. Spraying water before the compressor wheel may cause damage. But this will almost certainly be best for atomisation. So how to Prevent damage.. I started to research air flow through a static blade into a rotating blade. This is very interesting. I will not bore you with detail and simply explain its been tested and the results are good. 

 

Pre swirl.png

 

This is best described as a pre swirl. This will be placed directly before the compressor wheel. This is fixed and does not rotate. Water injection will hit this before the compressor wheel. Maybe this will better protect the wheel. To show the advantage of this the following pictures show flow through a turbo with and without this in place.

 

111.png

 

Above is with pre swirl

 

Below is without

 

2222.png

 

I have read lots on this and the trials and tests on turbo improved air flow and reduced vibration, load. Sounds to good really to not at least try it. This is incorporated into the air inlet 

 

A few threads to read, if they help.

http://www.methanol-...formation_id=54

 

https://themotorhood...-meth-injection

 

https://www.pistonhe...sp?h=0&t=680867

 

https://www.engineba... Injection.html

 

 

Other than the first videos I have all pages already bookmarked on the PC for reference : )

 

I've been researching this for some time and it does seem a good option for my build. I know what I'm doing is not the norm but if my cheap home made system shows any good improvement on charge temps and help hold back the detonation I will invest in correct parts. 

 

Thanks 

 

Rob