Forum » General Engine Building Discussion » Sodium filled Valves

Sodium filled Valves

General Engine Building Discussion

Talk about engine building here. New products, tricky questions or showcase your work - If it's engine building related it's welcome here.


Page 1
Author
140 Views

Does sodium filled valves make a huge difference on turbocharged engines?

I swapped Sodium Valves for Stainless Valves, due to not being able to get stock. I am having a knocking issue on that engine, causing me to retard timing and seems like the more I retard, the more heat I create and worse it becomes.

I do know the theory behind it, but asking from real world experience.

How do you know it's knocking? Are you hearing it knock or are you reading a knock sensor or some other electronic detection system?

Listening with headphones and seeing on knock detection on ECU.

2.6bar boost E95

Even at lower boost it is not taking as much timing its supposed to.

Cam timing within 0.5degrees where it is supposed to be, did double check that.

I retard the timing with 2 degrees, which in turn creates more heat, then I take away again and again and simply doesn’t stop knocking. Started with 15 and now on 3-5degrees.

Compression ratio, etc all stayed the same.

Can you give a bit more info on the build? What type of engine, turbo, ECU?

I'm amazed that you're seeing knock with only 3-5 degrees advance, however 2.6 bar is a quite a lot of boost (depending on the engine and turbo combination of course)

Cheers!

Have you tried spark plugs with a colder heat range? What are you using now?

Using 11 heat range NGK racing plugs, so that should not be a problem.

I unfortunately cannot give all the info as its partially undercover project, but I can say its 4.0 V6 twin GTX35

Motec M1

I did lots of tests like pressure loss across intercooler, etc. and could not come up with a solution. Also pressure tested all boost pipes for leaks.

The charge temp does go into 60’C, but with Ambient 30-40’C its not too bad.

Using 11 heat range NGK racing plugs, so that should not be a problem.

I unfortunately cannot give all the info as its partially undercover project, but I can say its 4.0 V6 twin GTX35

Motec M1

I did lots of tests like pressure loss across intercooler, etc. and could not come up with a solution. Also pressure tested all boost pipes for leaks.

The charge temp does go into 60’C, but with Ambient 30-40’C its not too bad.

I think you did lose quite a bit of heat capacity without the sodium filled valves. i found this article which says:

Sodium-filled hollow stem valves are available for higher heat applications, and are typically used for the exhaust valves. The sodium inside the valve stem melts and absorbs heat from the valve heat. As the valve opens and closes, the sodium sloshes up and down inside the valve to transfer heat from the valve head to the stem. This helps the head run cooler to reduce the risk of valve burning, preignition and detonation. The difference in cooling is quite dramatic. With a conventional solid stem exhaust valve, 75 percent of the cooling takes place across the valve seat and 25 percent through the stem. In a sodium-filled exhaust valve, 40 percent of the cooling is through the stem so the valve can tolerate more heat.

Yes, I am aware of the functionality of Sodium valves, this is more a question of practical experience?

would it cause the valve being that hot that the engine would be more prone to knock?

Other than that I am not sure why the engine would knock so badly.

have also swapped fuel and went up to E100, to rule out the fuel. Didn’t see any power difference between E95-E100, is that correct?

Probably nothing noticable. What sort of lambda are you targeting.

I can tell you that in my experience, on naturally aspirated engines you can hardly tell a difference between having and not having them; a degree or two of spark timing at most, which is why manufacturers sometimes eliminate them to save money. It certainly doesn't help that you got rid of them. 60C air temperature in the manifold during a WOT pull? That's enough to cause knock with that kind of boost. Ethanol % has a smaller and smaller effect on knock reduction as you increase it. So for example, E85 vs E100 is not a big difference, but E15 to E30 is a bigger change.

Another big cause of knock at high boost like that is high turbine inlet pressure due to the size of the turbine housing, the restriction in the exhaust post turbo, or the valve timing. I know you can't disclose all information, but have you tried phasing the intake and exhaust cams to see if you are at optimal valve timing? Not sure if you have fixed cams or not. Either way, the general rule is - retard intake cam and advance exhaust cam at high rpm & high boost. This lowers effective compression ratio and evacuates exhaust gases better, which is better for knock.

I've tuned more turbo race engines with stainless valves than with sodium filled and from my experience I don't believe the sodium valves make enough of a difference to create the sort of issue you're reporting. The advantage of the sodium filled valves would also be reduced slightly in your application given you're running E955 which runs cooler than gasoline. I'd agree with Raymond that the IAT is getting pretty high, even given your high ambient temps.

Without knowing your compression ratio it's hard to be too specific but on E95 it's usually difficult to even generate knock at lower boost levels - Admittedly 2.6 bar may well be enough but below perhaps 2 bar I'd expect you should be able to tune to MBT with relative immunity from any chance of knock. Certainly 3-5 degrees with the engine still knocking is not right!

The compression ratio is 9:1

Are you 100% certain you're actually getting knock? 9:1 is a low CR for dedicated E95 and even at 2.6 bar I'd be surprised if you were knock limited. If you were I'd be guessing your timing would still be easily into the high teens of low 20's. What audio knock system are you using? what frequency are you looking at in the M1 and can you post a screenshot of a ramp run on the knock control graph?