An In Depth Analysis of The MZR DISI Head

[Mazdazilla6]

Originally Posted by zenit Either you have very dull grinder tools, or you've put holes in whatever you're working on. 2 minutes of straight work in any once place is a lot of material removed if you're using the right tools.

I don't have Al specific carbides. And it takes a good amount of time to remove the intake port divider all the way down the runner. So I guess you could say I put a hole in it lol. My carbides cut pretty fast but even still, it's a lot of material to remove. Not only that but even moreso, I spend a ton more time with cartridge rolls than I do the carbides.

[Fstrnyou]

Originally Posted by Mazdazilla6 Interesting, my 3hp 12gal couldn't keep up with my die grinder at all. Literally grind for 2 minutes and it's spraying water out of the grinder from condensation. I was going to get a moisture separator but decided against it because the motor gets hot as hell anyway and I don't want to kill it.

Sounds like you need a larger tank...and to drain your condensate more often. Most compressors are limited by the pump, not the motor. And most single stage pumps should be kept under 50% duty cycle (1/2 time on, 1/2 time off). A larger tank would lengthen the off time of the compressor. If your running the tool continuously for extended periods, the pump CFM should be higher than the CFM of the tool, otherwise the tank will never refill and keep running until you stop the tool and give it a chance to catch up.

With yours only being 12gal, I'm betting you have a direct drive oil free pump. Those pumps run stupid hot and might even call for a very low duty cycle in the manual, but I don't have any professional experience with direct drive.

That's the long winded version of "subbing".

[Fstrnyou]

Any head flow testing in the works with the ST manifold?

[Mazdazilla6]

Originally Posted by Fstrnyou Any head flow testing in the works with the ST manifold?

Yea I tested that manifold a while ago. the manifold outflows the head by a good amount. Flow distribution is within 3% for pretty much the whole lift duration. I'm pretty sure the results were posted by DM in one of their threads.

[WetzMS3]

#stmanifoldsforharambe

Sent from my XT1254 using Tapatalk

[Fstrnyou]

Originally Posted by Mazdazilla6 Yea I tested that manifold a while ago. the manifold outflows the head by a good amount. Flow distribution is within 3% for pretty much the whole lift duration. I'm pretty sure the results were posted by DM in one of their threads.

My mistake. I've seen that thread but either overlooked the inclusive head test or forgot about it. I was thinking the flow comparisons were for the manifolds only.

Carry on...

[Mazdazilla6]

Finally getting back into this. Finishing up a port job on one cylinder soon and will hopefully be flowing it next week. Port dimensions aren't changed much, just cut down the divider (only about 2.5" long now) and matched up the small port to the large port. Just need to blend and smooth evergthing and if I like the results I'll be doing this to the head I'm going to use.

[Enki]

I'd leave the runners as they are to see how much flow gain you get from runner size matching.

[Mazdazilla6]

Originally Posted by Enki I'd leave the runners as they are to see how much flow gain you get from runner size matching.

Thought about it but I don't feel like taking the time to do another test port. Our airflow expert at work examined my head for a bit and gave me some advice. He's been porting and flowing heads for longer than I've been alive so I trust his opinion. One of the things he told me is every time he's messed with valve dividers that shortening them has always picked up a good amount of flow.

[Voltwings]

Originally Posted by Mazdazilla6 Thought about it but I don't feel like taking the time to do another test port. Our airflow expert at work examined my head for a bit and gave me some advice. He's been porting and flowing heads for longer than I've been alive so I trust his opinion. One of the things he told me is every time he's messed with valve dividers that shortening them has always picked up a good amount of flow.

For the sake of discussion, and probably more just my own understanding, would NA vs boosted have an impact on this at all?

[Mazdazilla6]

Originally Posted by Voltwings For the sake of discussion, and probably more just my own understanding, would NA vs boosted have an impact on this at all?

Define this.

If you're referring to airflow in general, I would guess yes. I've never seen someone flow a cylinder head with highly pressurized air. At the end of the day, for the most part, anything you do to improve flow on an NA application will be just as beneficial if not more on a boosted application.

[Voltwings]

I think you answered my question. I was thinking NA and boosted applications would have different flow characteristics (with regards to the head divider), but like you mentioned; at the end of the day, all that really matters is flow.

[Enki]

They don't really have different flow characteristics, but boosted engines can increase the pressure to pack more air into the cylinder; that's why there's two ways (that I've seen people use anyways) to report volumetric efficiency of a boosted engine.

You can say that both on and off boost (14.7 PSI) an engine has 100% VE (for example) or you can say that an engine has 200% VE @ 14.7 PSI. It is my understanding that both are technically correct.

Headwork and other flow mods allow for more airflow (power) at lower boost levels. A good example of this is disabling VVT on a stock turbo car; boost will go up in the lower RPMs but power (torque) will be down. If memory serves, with no VVT on a stock turbo, I couldn't keep the boost under 20 PSI below ~5k RPM.

Food for thought.

[Mazdazilla6]

Spent some time this afternoon finishing up the test port. Gonna make fixturing this week to get it set up on the flow bench.

[zenit]

It's amazing what a lil' spitnpolish'll do.

[Mazdazilla6]

Originally Posted by zenit It's amazing what a lil' spitnpolish'll do.

Our airflow expert at work gave me tons of pointers. I'm gonna try and start doing the exhaust side after I get the intake bench marked and tested. Basic plans are to open up the top of port at the flange and narrow the valve divider to open up the port without increasing its area.

[Mack_Sant]

Originally Posted by Mazdazilla6 Our airflow expert at work gave me tons of pointers. I'm gonna try and start doing the exhaust side after I get the intake bench marked and tested. Basic plans are to open up the top of port at the flange and narrow the valve divider to open up the port without increasing its area.

I have the old head from my old engine sitting around and I was itching to do something. Then the thought popped into my head to port it as practice for the engine that's actually in my car. Same day I found this thread. Can't wait to see the exhaust side.


Sent from my iPhone using Tapatalk

[Mazdazilla6]

Originally Posted by Mack_Sant I have the old head from my old engine sitting around and I was itching to do something. Then the thought popped into my head to port it as practice for the engine that's actually in my car. Same day I found this thread. Can't wait to see the exhaust side. Sent from my iPhone using Tapatalk

The head I'm using is junk. If your spare head is useable it might be a good idea to practice on a cheap head from a junkyard or something before trying your hand at a head that's worth something. It's a good learning experience.

[Mack_Sant]

Originally Posted by Mazdazilla6 The head I'm using is junk. If your spare head is useable it might be a good idea to practice on a cheap head from a junkyard or something before trying your hand at a head that's worth something. It's a good learning experience.

Oh it's junk. Old engine zzb. Cylinder 3 valves had an orgy.


Sent from my iPhone using Tapatalk

[acidplant]

Great work Mazdazilla6........ I highly commend you for putting great info to this forum. Knowledge is always needed and welcome

[Yatta]

looks purty, are you concerned about swirl?

[Mazdazilla6]

Originally Posted by Yatta looks purty, are you concerned about swirl?

No, the tangential port design should still provide pretty good swirl.

After I get it baselined I'll be throwing it through the bandsaw to see where all the water jackets are.

[Yatta]

Originally Posted by Mazdazilla6 No, the tangential port design should still provide pretty good swirl. After I get it baselined I'll be throwing it through the bandsaw to see where all the water jackets are.

up down versus side to side.. I understand that, just curious; can’t wait to see the cross section!

[zenit]

Isn't the need for combustion chamber swirl not as important for direct injected engines?

[Enki]

Originally Posted by Mazdazilla6 After I get it baselined I'll be throwing it through the bandsaw to see where all the water jackets are.

THANK YOU!

Originally Posted by zenit Isn't the need for combustion chamber swirl not as important for direct injected engines?

Actually it's probably more important, due to fuel being injected directly into the cylinder. Hard to get a good mix going on if shit just sits there eh?

[Mazdazilla6]

Originally Posted by zenit Isn't the need for combustion chamber swirl not as important for direct injected engines?

In any engine, you want as much mixture motion as you can get to a point (look up Reynolds number). My guess would be that while having more mixture motion would be beneficial, it's not as important on a DI engine because fuel pressure is so high that the fuel atomizes pretty good from the getgo and that might help it homogenize a bit better. But that doesn't mean you have a uniform mixture, so in that regard I'd also agree with Enki.

Whether it's more important or not, you shouldn't sacrifice flow for mixture motion in the port unless it causes you to lose all mixture motion. In our case, there isn't much to be lost by porting the head because 4 valve heads naturally make good tumble and our offset port design helps create swirl.

[Enki]

Originally Posted by Mazdazilla6 In our case, there isn't much to be lost by porting the head because 4 valve heads naturally make good tumble and our offset port design helps create swirl.

Another reason I left the runners in my head stock.

[Yatta]

Originally Posted by Enki Another reason I left the runners in my head stock.

I always thought the disparity in size was solely for this, there is enough material there for matched and Mazda clearly gave a lot of thought to the head (both for performance and economy, not necessarily fuel economy but that helps)... even with VTCS removed there is induced swirl... excluding the dumb-ass bump on #4 I think it is a good design. The Intake manifold on the other hand...

[Mazdazilla6]

Finally got around to working on the exhaust side. Didn't do a whole ton, narrowed up the valve divider and blended it in a bit better. Left the ceiling where it is and the opening at the head flange where it is to see what gains are there after I benchmark it. I've got all the fixturing made up to get the head on the bench but unfortunately I'm stuck waiting for my boss to finish flowing his dirt bike head before I can use the bench

[Mazdazilla6]

Single runner w/bowl work vs stock runner results. This time they were flowed at 28" so the numbers are higher for that reason.

image.png

Graphical form:

Capture.PNG

[Enki]

How much of that do you think is just from the bowl work?

[Mazdazilla6]

Originally Posted by Enki How much of that do you think is just from the bowl work?

Kind of hard to say. The reason I say that is that the large port side I didn't really alter, so it didn't pick much up in flow. I tested that my flowing one valve at a time. The small side on the other hand got a bunch of bowl work but it also got a fair amount of runner work. My thoughts/observations lead me to believe that bowl work will give a certain amount of gains before the small side becomes a restriction. I think this because after flowing each valve by itself, the large side (despite being a single runner) was still flowing ~15% more air than the small side, that could be contributed the the smaller cross section that's still before the valve on the small side.

Another thing to note, I messed around with probing the ports with a string/flag and that small port is almost completely dead on the upper left ceiling at the entrance. The single runner did pick up a bit more motion there, but not much. The large reason for that is the JMF design. The angle that the manifold meets the head at is simply atrocious. It causes the floor to become extremely active and reduces the flow on the ceiling a good amount. The design can be improved by angle milling the manifold flange or streamlining the runner better by creating a half teardrop shape on the floor of the manifold runner. Or just running a manifold with the proper geometry *cough* FoST *cough*

I'm going to play around with this head more and see if there are any good gains to be had in the .150"-.250" area. I'm pretty happy with the peak gains and I don't plan on running more than a .450" lift cam so this suits my needs fairly well, I just need to see if there's more under the curve.

Edit: Just realized I never added fixture setup photos. I made a radiused entry inlet to simulate the runner of a JMF since that is the manifold I'm running. Using an entry with an angle that agrees with the head would definitely increase flow.

[Enki]

I would have thought that the FoST mani would be just as bad...You're saying it's not? If so, this makes me happy.

[zenit]

Originally Posted by Mazdazilla6 I'm going to play around with this head more and see if there are any good gains to be had in the .150"-.250" area. I'm pretty happy with the peak gains and I don't plan on running more than a .450" lift cam so this suits my needs fairly well, I just need to see if there's more under the curve.

I'm jealous of your resources.

I don't think anyone makes a cam with more than ~.450" lift, or am I missing something?

random question- reading back through old posts @06Speed6; mentions combustion chamber work, which I assume is to deshroud the valves. thoughts?


also, @Mazdazilla6; if I buy you a couple +1mm intake valves, would you be willing to throw them on the test head and flow numbers?
I assume your career racecar shop has a valve seat cutter or something similar. I will buy you much beer.

[Mazdazilla6]

Originally Posted by zenit I'm jealous of your resources. I don't think anyone makes a cam with more than ~.450" lift, or am I missing something? random question- reading back through old posts @06Speed6; mentions combustion chamber work, which I assume is to deshroud the valves. thoughts? also, @Mazdazilla6; if I buy you a couple +1mm intake valves, would you be willing to throw them on the test head and flow numbers? I assume your career racecar shop has a valve seat cutter or something similar. I will buy you much beer.

Piper makes a cam with .462". We can't go much further than that before the retainer hits the seal.

CC work can mean a lot of different things. Deshrouding valves, removing potential hot spots, raising or lowering compression, altering squish zones. As for deshrouding valves, there isn't a whole lot that can be done on our head. There isn't anything that can be done between the valves and the bigger they get the smaller that clearance gets. There's a bit of material that can be taken off on the outsides but the biggest gains in terms of deshrouding the valves would be moving to a larger bore.

Gonna have to pass on that. We have an extremely nice Serdi seat cutter but I'm not trained on it and don't have the time to learn. All of our engine builders are indisposed as well with some new race program stuff.

I wouldn't be concerned with larger valves, the amount of work it takes to cut larger seats and the cost of the valves makes it not worth it at all. The exhaust side would probably benefit from +1 but again, the cost is pretty high and our stock valves are higher quality pieces than most of the aftermarket stuff IMO. Not to mention you've gotta change your guides to run aftermarket valves.

I will say though, I think there would be some pretty decent gains moving from the stock 3 angle cut to a 5 angle cut. It'd definitely help more than moving to larger valves.

[Enki]

Looks like for a stock head there's not much benefit to going to larger lift cams; obviously duration would be better but lift doesn't appear to give much back.

While I have bowl work on my head (intake side only), my cams only go out to .304 lift intake and .322 lift exhaust so I'll be bearing that theory out at some point.

[Mazdazilla6]

Originally Posted by Enki Looks like for a stock head there's not much benefit to going to larger lift cams; obviously duration would be better but lift doesn't appear to give much back. While I have bowl work on my head (intake side only), my cams only go out to .304 lift intake and .322 lift exhaust so I'll be bearing that theory out at some point.

I've been telling people for the past year that cams won't do much on a stock head. Anything more aggressive than the CS grind is pointless. Duration helps, we can ddfinitely get good gains from it.

[Enki]

Hey you wouldn't happen to have any information on the exhaust flow for a stock and/or gasket matched head would you?

[Mazdazilla6]

Not yet. Might stop in this weekend to do some flow work if I'm feeling good enough. If so I'll probably get to it then.

I've got a lot to say about the exhaust side because it's an interesting design but I'll wait until I've got numbers as well. Don't want to go making wrong assumptions before I have some empirical data.

[Enki]

That would be great. I just plugged the prior posted flow numbers into the engine sim and gained a little bit of power. Adding lift gains a little bit more, but not that much compared to what the duration on the cams got me.

Intake airflow:


304 lift (installed cams):


419 lift (same duration same angle higher lift):


Obviously this is only applicable to my build but still, adding an additional .115 (37.8%) lift isn't worth the extra 16 chp and 4 ctq I'd get.