PatrickT

Im considering making a new intake for many years, even bought a AC/DC TIG welder because of this 7 years ago.
The runners on the stock TII intake may not suit to my bridgeport anymore because timings are much different.
So I tried to calculate the length for the runners for many years, collecting patents Mazda made back in the days.
There seem to be different effects that create resonace:

1. what I learned in university:, and what mazdas tech paper for the 26B intake seems to depends on
.Intake (valve/port) opens thus creating a underpressure wave going up the runner.
.In the plenum the underpressure wave gets thrown back as a pressure wave
.Pressure wave goes down the runner in the combustion camber just before the intake closes.

- Timing for the wave is depending on the port opening time
-Plenum needed, or ITB in free air
Formula is:
runner length (in m) = port opening duration /360° * 60/RPM * speed of sound / order of resonance / 2

Mazda used 3rd order and 343 m/s as speed of sound for the NA 26B
Port opening duration in a PP is 360°

For 8500rpm I get 40cm, and for 6000rpm I get 57cm
with a difference of 17cm it looks like it fits to Mazdas diagramm in the tech paper




2.what all pistion engine calulations I found depend on:
-Intake closes -> air flow hits closed valve -> pressure wave generated
- pressure wave goes up the runner
-in plenum gets thrown back as underpressure wave
-underpressure wave hits still closed valve and gets reflected as underpressure wave
-underpressure wave goes up the runner
-in the plenum it gets reflected as pressure wave
-pressure wave hits the valve just as it opens

- timing depends on closed valve duration
-plenum needed or ITB in free air again
-you need a multiple of second order resonance because the wave has to travel the runner twice up and down to get a positive pressure wave

Formula:
runner length (m)= ( cycle duration - valve closed duration )/cycle duration *60/RPM * speed of sound / order of resonance / 2
cycle duration on a piston engine is 720°, on a rotary it would be 1080° till the same camber passes the intake again, but there a 3 cambers so it's 360°
valve closed duration on a rotary is very short like 72° stock and on a bridgeport or PP none existend

For 7000rpms and 2nd order I get on the stock engine a runner length of 11cm, that would be barely more than the length in the side housings...
Higher order will only shorten the needed runner length, also like porting
IMO not appliable on a rotary


3. what all Mazda patents for street cars depend on
-when intake opens there is still a higher pressure because of exhaust gases in the combustion camber -> creating a pressure wave up the runner
- pressure wave travels from one runner through the manifold down the runner of a different rotor
-there the pressure wave goes in the other combustion camber just before the intake there closes

-timing depends on exhaust pulse ( intake opening ) and intake closing timing
-manifold needed ( plenum not needed )
-I think works best on 2 rotors

Formula:
total length = ( intake open duration -180° - delay ) /360° * 60/RPM * speed of sound

total length means: length of runner 1 + offset between the two runners in the manifold + length of runner 2
as delay Mazda uses 20° ( maybe this is also appliable on the other fomulas )

Foir the stock TII engine, I get about 110cm total length, at 5000rpm and a speed of sound of 380m/s because of boost pressure/temperature
I think this matches the TII runners with 45cm each with a offset of 20cm in the manifold...


But !! for a bridgeport like mine, the intake is already open when the exhaust opens on the next camber.. so I believe the pulse starts right when the exhaust opens
...to be continued...

PatrickT

So today is pictures day:

About the first point where I try to reverse engineer the 26B intake:
I took the drawing from the 26B and measured the intake in relative to the rotor housing inner height, which should be 234mm ( please correct me if im wrong)
So the intake is in the short position about 400mm long

In the paper it says this short position is at 8500rpm, while at 6000rpms it is 175mm longer -> so about 575mm

Can I upload the whole pdf here?

This is what I get by this calculations for the stock TII ports and runners:

In the top row it calculates the length for a given rpm and in the bottom the rpm at which resonance will occur at the given runner length.
I don't know at which rpm to look at for the stock engine, but in the bottom it says the runner / port combination does resonace at 6000, 4800, 4000 and 3400rpms
Sounds legit, but I can't prove this right now. I think the engine has highest torque at about 3400-3500rpms

For the 26B I just put intake opens and closes at the same timing, so they are 360° apart.
And for 8500rpm I get the runner lengths of
1st order 1,21m
2nd order 0,61m
3rd order 0,4m ( which is what I assume the given length)
and so on...
Below with the given length of 0,4m ( which is 400mm )
We have resonance at 25700 , 12900 , 8500 , 6500 rpms etc..



I hope this is right, if it is not, shame on me


Now to the 3. pointm the exhaust pulse that charges the other rotors intake.

I have made a CAD model of a 13B housing with rotor and stock ports.
The rotors has marks on it where to read the "crank angle"
"OT" means "top dead center" in german, and the dots from there are in 10° steps.

So this ist the rotor at top dead center for chamber 1:

Exhaust is still open in chamber 1. Chamber 2 is comressing and chamber 3 expanding.
The chambers timing is 360° apart from each other.

So here chamber 3 is opening the exhaust, high pressure gases evade and because of the turbo, a pressure builds up with rising rpm.
The pressure also goes into chamber 1 , where the exhaust is still open.
While the intake of chamber 1 is still closed...



Now the intake in chamber 1 opens , and what Mazda explains in their patents, the exhaust pressure still traped in chamber 1 evades in the intake runner.
This only occures when the exhaust pressure is higher than boost. Which happens due to their patent from 5000 rpms on.

This is very bad for the filling of course, and to neutralize this, they want to use the pulse to charge a chamber on the other rotor.


This is the position of the other rotor in a 2 rotor engine, it's 180° apart.
You see chamber 1 is there about to close. So the pulse has got this remaining time to pass from one rotor to the other, to enter there right when the intake closes.



So in the calculations I made an other part markt "X" for "cross-resonace" whatever.
And the port timings of the "receiving chamber" is 180° delayed.
This corresponds to the Mazda patent with their formula, and I get 1,11m total length




But what does it mean for a engine with a bridge or PP ?
To be continued...

TeamRX8

iTrader: (2)

kind of confusing because you seem to be talking about and intermixing both NA and FI/Turbo; what are you actually doing this for?
.

PatrickT

I am doing this for a BP turbo.
It is also sometimes confusing to me but as far as I understand the rules are the same for NA and turbo.
The NAs also seems to have residual exhaust gases in the chamber, maybe from back pressure from the exhaust system.
There are Mazda patents from the 80s for this exhaust pulse effect for several engines,
A 2 rotor in general , with different intake closing timings, with aux ports and with turbo.

TeamRX8

iTrader: (2)

Back in the day of slow responding, less efficient turbos maybe …
.

PatrickT

This is the patent I am refering to, and I believe the TII intake is designed on this:
Intake system for rotary piston engine; patent #4614173

They say for this "exhaust pulse cross resonance effect" to work, the exhaust pressure has to be larger than boost by more than 100mmHg. Which is 0,13bar or 2 psi.



This is a diagram from the patent, showing the boost (Pin) versus the exhaust presure (Pex) over rpms:


While the boost is controlled by the wastegate to stay constant, the exhaust pressure raises with rpm.
So exhaust pressure gets larger than boost.
This is because of stock turbine sizing and the very restrictive stock exhaust system.
I believe with a big turbo and free exhaust this effect is gone on the stock ports.

Spider2k

So slightly off topic but close.

Will an FD upper intake manifold mated to an s4 NA 6 port lower work? I can't find a way to mount a DBW throttle on the NA. All the adapters are for the FD and Cosmo uppers.

Manny_Apex

iTrader: (1)

Very off topic; start a new thread. The people who would open and read this thread (ie me) may not be the same people who would be able to answer your question, and you're also in the wrong section. Post it here:

2nd Generation Specific (1986-1992) - RX7Club.com - Mazda RX7 Forum

Don't assume you're the only person with this question because you're probably not. By making a new thread, you'd be helping more people than you will probably ever know.

TeamRX8

iTrader: (2)

I get all that, but the reference is the technology and understanding of the time; the early 1980’s.

.

PatrickT

Of course, first I want to understand what they did and for this I want to reverse engineer the TII manifold.
Maybe I can use the principle on my project. The physics behind this didnt change.

I'll bring updates soon, I was just to busy these days.
But it looks like I did the measures of the TII runner lengths wrong when I first did them many years ago...

PatrickT

So I started 2 weeks ago taking the measures of the ports on the engine block.
I had them slightly ported when I build the engine.
Primaries are: 38mm in heigth and 29mm in width -> which is 645mm² in area
Secondaries are: 41mm in heigth and 29mm in width -> which is 1008mm² in area




First I thought about building a new upper and lower manifold, because I wanted to get rid of all the stuff I dont need
and wanted to adjust the runner length and diameter to the ports.

I measured the inner diameter at the flange to the uim:
32mm for the primaries which is 803mm² in area
and 38,5mm for the secondaries which is 1163mm² in area.

Both areas are already larger at the flange to the UIM than to the engine.
So I thought there is no need to build a bigger LIM.
I can also leave the secondary injectors there and dont need to deal with builid sockets for them etc


So for building a new UIM I want to adjust the runner lenghts.
First I want to know the stock runner lengths.

Runner length in the intermediate housings is about 65mm:


In the side housings about 75mm:


In the LIM the primaries are about 240mm in length,
and the secondaries about 250mm:





Now the difficult part:
In the UIM I used an old belt to lay it in the runners to let it just lurk into the plenum by about half the runner diameter.
( This is where a uder pressure wave gets thrown back as pressure wave )

For the primaries this is only about 90mm

For the secondaries its about 110mm



This means, the total runner lengths from combustuion chamber to plenum are about:
Primaries: 395mm
Secondaries: 435mm

On the TII the port timings are the same for primaries and secondaries ,
Port opending at 32° after top dead center and port closing at 50° after bottom dead center, which results in port opening duration of 288°

With the formula:
RPM = ( port_opening_duration - 20°) / 360° *60 / runner_length * speed_of_sound / 2 / resonance_order
(-20° because it takes a time to form the pressure wave )
So for the secondaries I get resonances at 6500 , 5200 , 3500
Because of the raising residual pressure with rasing rpm this effect may get negativ at high rpms , so only the one at 3500 may give some benefit in stock configuration.

For the primaries I get resonances at 7200 , 5800 , 3800 but I must say they barely have a plenum volume and may have no big effect.


So next to the cross resonance:
For this I measured the route from one primary port through the "primary plenum" to the exit of the other primary port:
Same for secondaries..
I got 405mm for the primaries and 505mm for the secondaries:




To total length for primary "cross resonance" is :
65mm (intermediate housing) + 240mm (LIM) + 405mm (UIM) + 240mm (LIM) + 65mm (int. housing) = 1015mm = 1,015m

Secondaries:
75 mm + 250 mm + 505mm + 250mm + 75mm = 1155mm =1,155m

So with mazdas formula from their patent solved for RPM:

RPM = ( port_opening_duration -180° -20° ) / 360° * 60 / total_length * speed_of_sound
( -180° because the rotors are 180° apart in a 2 rotor, for a 3 rotor it is 120° )

For primaries I get 5500rpms
and for seconfdaries 4800rpm, which is oddly low, while mazda is always refering to a rpm range of 5000-7000 rpms

Maybe it is in average just about around 5000rpm, which I thought in the past it is the calulated rpm.

Or maybe someone finds a mistake, so we can correct it ?


So to the offtopic questions:
If someone could measure the runners in a FD UIM .... it would be very interesting and might help reverse engineering everything

-Patrick

TeamRX8

iTrader: (2)

the physics didn’t change, but nobody could even imagine back then having a turbo manifold design that created internal expansion to effectively intercool the pressurized airstream inside the manifold like exists on the Porsche engine now. Most people today still have trouble grasping it even. Back then they also didn’t have efficient turbos capable of providing over 1.5 bar boost under 3000 rpm in a wankel either. Therefore the design parameters for that are different now.

not trying to be difficult or harass you so this is my last post about it, but if you’re going to the trouble to design and produce a manifold then doesn’t it make sense to built it to 21st century standards rather than 20th century? Any way, best wishes.
.

PatrickT

Yeah I thought about that, in the GT2 they compansate for that with a higher boost level.
I dont know if this is the right aproach on a BP with a huge overlap, but maybe it is fine.
Also the 911 GT2 has troubles to properly intercool, which I have not.
After all that I already came to the conclution that I wont go for a plenum at all, but mainly aim for a good flow.

Just make the runner lengths that way that a resonance wont hurt but help if one occurs.
And I just wanted to show how I came to the numbers.

So little jump ahead.
This is what my current design looks like:



And this is a first flow calculation:

Maybe I can get rid of the red and green spots and make it all yellow

-Patrick

j9fd3s

iTrader: (3)

conventional wisdom for a high overlap engine is to keep the ports separate, its thought to help with keeping each rotors gasses separate, you're not mixing the exhaust gas from one rotor into the other
also granted there have been plenty of bridgeport turbo FC's and FD's... again the 1990's way is like your paper, if the backpressure is more than 2psi higher than the intake, you loose.

gracer7-rx7

iTrader: (16)

Defined autoworks posted some good old threads on intake design ages ago. Not for a turbo power plant but interesting reads. Various other people have posts on the topic as well

fyi

PatrickT

Exactly, this is why I chose a EFR 8374 to have no backpressure till 6500, asuming Borgwarners matchbot is right.
So my runners are going to be separate for 52cm , which is a litte more than stock primaries, so it should be fine.
I also think worst thing you can do is like removing the "wall" between the primaries..

This 52cm is the runner length where they join, so there is no additional length from one rotor to the other, so total length goes down to 104cm
This should move the exhaust charge effect from 6200 ( where there is none with the EFR ) to 6750
and the runner harmonics from 8850 down to 6900 ( stock runner length is 44cm )

...welding in progress

PatrickT

So I already made the flange weeks ago and then cut 2 180° tubes to 90° each.
45mm for the secondaries and 38mm for the primaries.
Made the 45mm and 38mm merge in a 10° angle and form a shape with the same circumference like a 60mm tube:


Than having a 60mm 180° tube cut down so it fits the both holes and to join in the middle.
Squezzed the 60mm tubes in a vice, hammered and bent them to fit these pear shaped tubes.


Welded it together.


Cutting the front open again and extending it with 60mm tubes-sides to fit the final 76mm = 3 inch tube:

It is always like, cutting it to shape, bending it to make smooth transitions,
welding it from outside (picture above), welding it from inside and grinding it from inside to make it smooth already.
Because later on I won't be able to reach the inside again.

This is it so far: the 76mm tube is just squeezed in the vice down to 60mm to fit the opening of the picture above,
and it had to get a like 10° kink to align with my intercooler piping:

Looks kinda wrong as an intake but let's see how it works XD
Now it just needs the throttle body in the straight 76mm part , some vacuum pipes, and finally some finishing and it should be done.

Greets, Patrick

PatrickT

Flange of the throttle body is now in place.
Welded from outside and inside and grinded smooth again.
Grinded the flanges on a stone board with sand paper on it.
It's a poor mans face milling XD

Next I'm gonna ad a 10mm vacuum socket for the brake booster, and a 6mm for all the other vacuum lines

PatrickT

It's done!
Finally welded the vacuum sockets and the flange for the throttle body to the manifold


I'm using the stock air intake temperaure sensor right in front of the 70mm throttle.


And thats it on the engine:


Now I only need a 76mm to 70mm reduction hose to connect it to the intercooler piping

HarryW

Not bad,thanks for the photo

PatrickT

First started it yesterday, looks like it has a less tendecy to brap.
70mm throttle is more sensitive than the stock...which was to be expected..
And yeah, I need a new 80mm clamp XD

fidelity101

iTrader: (11)

Dang I should have consulted you before building my own intake. I used some very basic math and assumptions and went for it but the science behind it I am a bit iffy on.

fdracer

Please don't take this as shitty criticism, but may I ask why you didn't get a pro to weld it up for you? You put all this thought and effort into the build; it's a bit of a waste for it not to be beautiful. If trying to save money, you could've just tacked it up and had a pro do the final welding to make a it a work of art.

PatrickT

Tbh I also just made assumptions that were unproven.
But today I was at the dyno, doing the mapping.
Target was 400hp and when we reached that goal at 7000 and 1bar boost ( 14psi) I called it a day.

By looking at the graph it looks like it has some resonace from 6000 to 7000.
Which I think was how I intended it to be XD

I hoped of getting the 400s with less boost, but I saw later at home it whould make a lot more power with more rpms.
And we had quite a bit of slippage on the dyno and I dont know if they take that in account with the results...
(dyno says 6900 whlie my ECU says 7200)



Well I'm a DIY guy, I can say I built the engine completely myself. Like a lot of other parts on the car.
And its build to be a racecar. Opinions may differ, but I'd need an extra 20k to make it all look like a pretty car, but it wont make it faster.

TeamRX8

iTrader: (2)

Interesting. Your perseverance is both admiral and worthy of respect. Congratulations on accomplishing your goal.

The true test imo would be to replace the UIM with a properly sized open plenum having short port-entry trumpets and then see what the real difference is compared to theory. Because I'm still of the strong differing opinion that the much faster spool of the current turbo technology is going to influence the results differently than the testing results from way back in the day will suggest. They didn't have that opportunity to consider back then, they had to choose between quick spool followed by quickly ramping emap or the opposite. That there is a net resonance effect is neither doubted or questioned, but there's more to it now with other influencing factors not fully considered. And what it's going to come down to is how the magnitude of the resonance plays out against the magnitude of those other influences. Again considering what Porsche and others are accomplishing.
.