When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
For about 30 years I had a Paxton supercharged L98 with water injection before the throttle bodies.
I managed to hurt the pistons twice, always pistons one and three, and suspected the water was having trouble making the sharp turn into the forward runners.
Always wanted to visualize the flow and see.
Some interesting things.
With the dual throttle bores, the airflow from the bore closest to the runner blocks the flow from the other bore.
It gets deflected and has to turn around a come back.
Only runners six and eight are far enough to get equal flow from both sides.
Never realized that the throttle is so much lower than the runners.
The flow actually hits the bottom ramp of the plenum and turns up into the runners, especially the forward ones.
I think the side throttle option for the First Manifold would be a balance nightmare, looking at the different paths each runner set deals with.
I've always wondered why the runners mate with the plenum at a 90 degree angle and why they aren't pointed toward the throttle body. Seems like it might flow better.
Very cool. I sure wish I was smart enough to interpret it...
Question regarding the wet flow and the turn to runners #1 and #3: Would a larger throttle body slow down the air enough that it could easier navigate the turn into 1 and 3? (Can you cut a big single hole in your balsa wood "throttle body" for giggles?)
How would the airflow change if you simulated a large single throttle body (like the FIRST intake has)? -What if you removed the EGR bump / flattened it out -it's odd that the FIRST includes the EGR bump on the bottom back of the throttle body even though they don't actually machine in functional EGR. Some folks have a theory that it is there to help even out air distribution at part throttle -is that "testable" on the bench like this?
-The airflow to cylinders 6 & 8 get all crazy turbulent from the "box" in the rear passenger side of the plenum -if that gets smoothed out or ground down (it's solid cast in the FIRST; not sure in stock TPI), does that at all reduce the turbulence to 6 & 8?
What if the rear wall of the plenum is straightened out like you cut out the back and just had an angle straight across the back of the plenum? (Always wondered how bad that S-curve rear wall is and with a CoP conversion and a vortec truck cam sync sensor you don't need the space for a distributor and can do whatever you want with the back wall.)
Adam
Last edited by newbvetteguy; 01-20-2022 at 06:15 PM.
I saw on your YouTube channel you posted siamesed runner and monoblade throttlebody tests!
Am I just seeing what I wanted to see or does it seem that there is some reduced turbulence on average vs. the dual throttle body tests?
Adam
I think I agree generally...
It looks to me like the Siamese ports make it easier for the flow to make the turn into the runner pairs, and that may be a benefit of sharing the runners on top of the increased area.
The Mono-Blade does appear to have cleaner flow patterns and turns better into the runners also.
I'm in the process of adding another pump and recalibrating my flow bench, but I did flow the Edelbrock base, AS&M runners and 52 mm throttle combination.
Haven't quite matched the numbers up to what I see in the videos yet?
The lower flow in runner 1 makes sense.
Maybe the low flow in runner 8 makes sense being boxed in the back corner, it can only draw from one side.
The uneven flow in the center runner pairs is a little mystifying?
The thing with TPI and long runners in general is the water hammer effect of the closing valves which sends pressure pulses up each runner into the plenum.
This effect cannot be modeled in a static flow or even dynamic condition without simulating the closing rates and frequency (different cams = different closing rates) of the intake valves....
It may have something to do with, OEM engineers may have designed the runners, positioning and pairing to coordinate with the firing order of the engine which is why we see a discrepancy in a static flow condition between runners on the bench
On the other what we see may be related to turbulence introduced by static flow condition, or even venturi effect of pressure drop localized near one runner (of a pair) which would cause a differential in pressure to the nearby runner since they are so close (causing the % difference in flow)
In other words, a real engine 'pulls' on one of the runners at a time. So watching them all flow simultaneous isn't relative or related to the real-world situation of each cylinder taking a breath individually.
I note possible relevance with the close-ness of 5 and 7 runners (traditional 18436572 firing order) as being potentially 'poor' situation for cylinder #7 right after 5 takes its breath. And then recall the fact that the LS firing order 1-8-7-2-6-5-4-3 , may be helpful or even aimed towards fixing this issue in all manifolds of the future.
Some stuff you can try
1. Try blocking off a runner here or there and witness whether it increases the flow of other runners
2. You can acquire 2-bar map sensors from Ebay ($12?) And use an arduino microcontroller ($20?) To monitor the pressure at the ends of each runner (use the full manifold perhaps for the full 28" Long runner or whatever it is length), this will allow you to data-log the pressure of each runner as you adjust the flow or diameters or whatever
#2 could be done on a running engine and the 2-bar will be able to detect the pressure increased caused by water hammer
Poor Cylinder 8... Isn't 8 already deemed the hottest cylinder on a typical cooling setup SBC? Maybe the reduced airflow help it run fatter and cooler??...
Regarding the siamesing helping to even out the air flow cylinder-to-cylinder and the uneven distribution, take a gander at a REAAAALY time consuming port job of a TPI intake (FIRST) that was designed to help even out the cylinder distribution in the 2nd post on here: https://www.corvetteforum.com/forums...p-kill-it.html
That IS NOT a small amount of porting effort...
I think Ezobens is a member of this forum, too... -I PMed him on here and the CF forum to see if he'd chime in on this thread. @1989GTATransAm Also has some experience with some modifications here, if I remember correctly he messed with the rear plenum wall in one of his epic builds...
Adam
Last edited by newbvetteguy; 01-21-2022 at 05:14 PM.
I note possible relevance with the close-ness of 5 and 7 runners (traditional 18436572 firing order) as being potentially 'poor' situation for cylinder #7 right after 5 takes its breath. And then recall the fact that the LS firing order 1-8-7-2-6-5-4-3 , may be helpful or even aimed towards fixing this issue in all manifolds of the future.
I've heard it stated before that the TPI intake has the potential to see the most potential gains from a swap to an LS firing order. It doesn't seem like 5->7 could be a worse idea on a TPI engine, so it makes SENSE.
(Usually that means nearly unmeasurable gains, except apparently as cubic inches get larger, but maybe it means small single digit gains on a TPI??? -Worth the cost increase other than for "Science"? Not so sure...
The thing with TPI and long runners in general is the water hammer effect of the closing valves which sends pressure pulses up each runner into the plenum.
But is there any design feature that can help attenuate the water hammer effect in the plenum?
I wonder if the simple "box" shape with both sides parallel to each other and the base of the plenum and the lid of the plenum being flat and parallel makes any waves coming out of a runner and into a plenum bounce around in a more direct way towards the other runners. Hood clearance ignored, would a domed lid on the plenum help to make it more random / attenuate back and forth waves coming out of one runner and bouncing back and forth from side to side? --Why do so many modern CFD-designed plenums have domed roofs, anyway?
I know the plenum is going to be pure chaos no matter what. There's a video showing a lexan plenum that John Kaase built and you get to see the just absolute chaos in the plenum as it runs the engine through the RPM on a dyno, then he pokes his finger through a runner while they do another pull and you can see his finger being thrown around by the waves and it gets so painful he has to pull his finger out before the run is over. -Such a great video showing some important things that happen in a TPI plenum and runners. FOUND IT:
But is there any design feature that can help attenuate the water hammer effect in the plenum?
iirc short runners and large tapered plenums (Look at the Greddy intake manifold for a SR20DET or RB25DET engine) may negate and shift the influence of majority of the water hammer, and disturbing physical features associated with fluid flowing around bends, so that it will work well at high RPM and high rate of flow (200hp/liter ranges are common).
Water hammer adds torque to some RPM Range +2% to maybe 6% VE increase is possible I would imagine, so OEM manufacturers like to gain some 5hp to 17hp or whatever through that range for natural aspiration
Most modern engines use water hammer for torque in some RPM range, for example 1zz Toyota and LS1 and TPI intake runner lengths come to mind. It is a desirable feature of modern engine design for stock engines (moderate RPM Ranges) which can be controlled and have limited RPM (specific frequency range used the most). TPI length targets 4500rpm around that. LS1 what, 5200? 1zz is like 5500 I bet.
The job of (any) engineer is account for, monitor, record, understand these type of fluid behaviors. Then sometimes, they can be controlled, or dealt with somehow. Vibrations can be very destructive and control theory is the branch of engineering concerned mostly with accounting for, modifying vibrations and signals which can be electrical or material in nature as matter is electrical in nature, interactions between atoms as covalent bonds is electron behavior
In a racing application, runner length design features can get in the way. For example trying to use an LS1 intake 5800+rpm is going to suffer the same way TPI does after 4800rpm, and most racing apps call for 7k 8k rpm instead so it is unwanted or seems that way to some. That is why the Greddy intake has a taper and very short thick runners, it can be used at any high RPM and likely tuned for high RPM as well with those short runners, but it contains no mid-range torque benefit which creates very low torque output from the 2L and 2.5L Engines in midranges. Therefore those manifolds eliminate the resonance benefits in typical driving ranges (We drive normal 3000-4500rpm usually?) allowing high RPM and powerful flow throughput but also making the engine a 'dog' (Slower) outside of high rpm boost conditions unless something else is done to help with that.
My advice from 20 years is focus on reliability and turbocharging. The turbo makes the intake manifold choice much less important, you can use almost any intake you want and get good results. Seal it up well, pressure test it, and go nuts
Although I understand what's trying to be said, the words Static and Flow have totally opposite meanings and probably shouldn't be used together.
Maybe constant or continuous flow would be clearer?
The Kaase video has been around and discussed a long time.
The flow in the runner with his finger does pulse with the valve motion, that's where harmonic/wave tuning comes from at IO and ram/inertia tuning comes from at IC.
I agree that visual smoke testing may have limited value here.
But as Kaase states the pulses are much weaker at the end of the runner, watching his finger motion also shows that.
We have all put our hands over a throttle while the engine is running and felt a constant flow.
Somewhere between the runner entry and the throttle opening to the atmosphere, the pulses dissipate and the flow in the plenum transitions to a constant flow.
I believe that happens quickly.
Early in the Kaase video you can see the air fuel flow in the plenum streaming from the carburetor to the runners, very similar to what we see in my smoke test videos.
Admittedly not perfect, I believe smoke testing constant flow gives a good visual of what is going on in the plenum.
We all know that the flow in the cylinder head and manifold runners is a pulsing flow, but we still test and learn about them on continuous flow benches.
Even the short duration pulses are subject to the same physical laws as continuous flow.
Sorry, When I said static I mean steady. Referencing a very specific condition in fluid mechanics with no derivative for velocity.
Page 84 Fluid mechanics "Flows can be either steady or unsteady. In a steady flow the velocity at a point on a streamline does not change with time.
The partial differential of
dV/dT = 0
This condition is useful in performance for reasons, because it becomes easy to target a specific steady flow rate and optimize for that rate.
We will have a much easier time testing the capability of an intercooler using steady flow.
The flow in a manifold can not be steady because of the opening and closing of valves always causes a velocity change, however the flow through an engine may be steady once it is isolated from the internal influences of the manifolds (upstream or downstream).
An engine that can maintain some constant rate of flow through some range of RPM will exhibit peak and hold power output, the power will peak to some number and stay as RPMS climb pretty much as a straight line of power as long as flow remains steady.
Example LS turbo steady flow from my car
When we look at a compressor map now, the engine from say 4000rpm to 7000rpm will just sit at some point on the map, since the x axis is flow rate and flow rate is constant.
Steady flow condition allows easily choosing turbo & intercooler selection for specific peak power and purpose (ex, drag racing where the engine quickly maintains some high RPM and never uses anything below 5500 for example).
As I already pointed out, The flow at the intake pipe and exhaust pipe may be steady while the flow inside a manifold is unsteady due to valve actuations.
Therefore when we reference steady flow with respect to an intake manifold (as I did but used the word "static" earlier, to clear up what I mean with that word) I am saying that while the rate of the fluid flowing through the manifold is steady or static (unchanging velocity on the pre-throttle body pipe as in a flow bench or peak and hold situation where the manifold is tested at some specific CFM for example 600CFM constant rate) The internal flow of the intake manifold has changing velocity components related to the opening and closing of intake valves.
This effect cannot be modeled in a staticsteady flow or even dynamic condition without simulating the closing rates and frequency (different cams = different closing rates) of the intake valves....
iirc ... and large tapered plenums (Look at the Greddy intake manifold for a SR20DET or RB25DET engine) may negate and shift the influence of majority of the water hammer, and disturbing physical features associated with fluid flowing around bends, so that it will work well at high RPM and high rate of flow (200hp/liter ranges are common).
I've noticed all the modern tapered plenum designs and wondered why... Thanks for sharing that tidbit.
I've seen it on 2JZ intakes and even the Holley Sniper Dual Plenum LS cross-ram intake plenum:
I believe a taper maintains air velocity for the incoming column of air. If the plenum is square shaped the pressure will be lower at one end, inconsistent across cylinders.
It would be A design feature which accounts for feeding cylinders as air flows towards the rear-ward cylinder, intended to keep them all similar VE (volumetric efficiency), cylinder fill
Wow. Yea it really is! CFD is so wonderful, when you can get it. (I'd LOVE to see a CFD animation of a TPI but I've literally never seen one, despite looking for a few years.)
I'm not sure that any of it matters for a TPI, because even if someone went crazy and built a modern "ideal' plenum for a TPI, they all seem to have the highest point at the very front of the plenum, right where you have the least hood clearance. Without chopping up a hood, it's kind of a non-starter.
It IS interesting to see the huge increase in velocity just on the plenum side of the throttle body in the top design vs at the tapered part of the runner in the 2nd one. And despite the higher peak velocity inside the plenum on the first one, you also have lower speed slow areas in the first design.
The top flow diagram shows the entry to the runners of tapered plenum design looking a lot more like a CFD diagram on a bell mouth inlet, too, while the square one doesn't look anything like it. Other than obviously still having vacuum in the plenum, the tapered plenum's airflow looks much closer to what I'd expect the airflow to look like with individual throttle bodies between the runner entrances and the plenum. -SO much less "plenum chaos".
-I'm not sure how much of a decrease in velocity is required to see fuel, water meth, or nitrous droplets fall out of solution, but those dark blue lower speed CFD arrows in the square plenum aren't going to help out any type of liquid injection prior to the throttle body, IMO...
Best engine-related post I've probably seen in a year on the internet. Thanks SuperZZ4!
Adam
Last edited by newbvetteguy; 01-25-2022 at 11:59 AM.
Just keep in mind those are static flow conditions, no closing valves. When a valve closes the momentum of air crashing into the valve changes the flow and velocity profile significantly, depending on RPM and flow rate. Also reversion due to long valve open periods while the piston is coming back up the bore (around low rpm) further has effects on the behavior.
Dynamic situation of running engine far more complex than basic static flow condition. Though I suspect when the engine is making full power there is a closer relationship with static flow conditions than say, mid-range torque where VE is high but flow rate is lower.
To visualize what I am saying think of the velocity of air once a valve closes, velocity must approach 0 and probably reverses many of it's direction components to gain velocity in a direction leading away from the intake valve and back up into the plenum
SuperZZ4, I'm assuming by now you've about had enough "fun" with the TPI intake on the flow bench.
I AM curious whether a spacer between the plenum and the throttle body would potentially help air make the turn into cylinder #1 easier. -Just like a carb spacer can help give air a straighter shot to its destination (and slightly increase plenum volume), wouldn't a spacer between the throttle body and the plenum do the same? (Or moving the front wall of the plenum forward more?)
SuperZZ4, I'm assuming by now you've about had enough "fun" with the TPI intake on the flow bench.
I AM curious whether a spacer between the plenum and the throttle body would potentially help air make the turn into cylinder #1 easier. -Just like a carb spacer can help give air a straighter shot to its destination (and slightly increase plenum volume), wouldn't a spacer between the throttle body and the plenum do the same? (Or moving the front wall of the plenum forward more?)
Adam
Adam ... this is all about learning to me and I'm never sick of learning.
My "keep busy" project this winter was to build a flow bench and I have a rather long thread going over in the Corvette Forum about it.
Problem I always have is going off in ten different directions testing things and also upgrading and improving the flow bench.
I had the same idea as you and figured I would replace the stock plenum with my model and flow test it on the bench, then test removing the rear vacuum fitting bump and add three inches to the front of the plenum.
I'm not getting the same flow numbers with the model as the real plenum.
The balance between runner is generally the same but the flow rates are higher and the two bad corners ( #1 and #8 ) aren't as bad.
Think it makes sense that they are different, the model is smooth inside and the real plenum is full of things impeding the airflow.
Removing the vacuum fitting bump only increased the flow in #6 and #8 about +5 cfm each.
The three inch spacer lost -4 cfm in runner #1 and gained +13 cfm on runner #3.
So when things don't make sense, I take a step back and lick my wounds.
Going to think about what's happening and try the tests again.
Maybe add a spacer to the real plenum and flow test that.
Took a break and finished machining some test and calibration disks to expand on the ones I bought from PTS for my flow bench.
Problem I always have is going off in ten different directions testing things and also upgrading and improving the flow bench.
I had the same idea as you and figured I would replace the stock plenum with my model and flow test it on the bench, then test removing the rear vacuum fitting bump and add three inches to the front of the plenum.
I'm not getting the same flow numbers with the model as the real plenum.
The balance between runner is generally the same but the flow rates are higher and the two bad corners ( #1 and #8 ) aren't as bad.
Think it makes sense that they are different, the model is smooth inside and the real plenum is full of things impeding the airflow.
Removing the vacuum fitting bump only increased the flow in #6 and #8 about +5 cfm each.
The three inch spacer lost -4 cfm in runner #1 and gained +13 cfm on runner #3.
So when things don't make sense, I take a step back and lick my wounds.
Going to think about what's happening and try the tests again.
Maybe add a spacer to the real plenum and flow test that.
Took a break and finished machining some test and calibration disks to expand on the ones I bought from PTS for my flow bench.
Great stuff!
I have ADHD so I understand the "novelty gene" and having a lot of unfinished projects all at the same time.
I did not realize that the OEM TPI plenums had so much junk in them... Things are much less busy in my FIRST plenum.
I'm shocked at your results building the 4" spacer and that the flow to #1 didn't improve by doing that...
I remember Wilson making comments that when they design a custom plenum (front facing throttle body EFI plenum) for a customer that they like to have a big "hump back" design that increases the area just after the throttle body and into the plenum to slow the air down so that it can make the turn into the first runners better.
I included pictures from my FIRST throttle body and the plenum, just to show how much less junk is in the way of the airflow in it vs. your triangle EGR passage era TPI plenum.
First throttle body with stock "EGR bump" at the rear (before I removed it) FIRST plenum: Less junk but still cast with the EGR bump down the middle and with the remnants of the vacuum box at the rear passenger side -I started with good intentions to remove the EGR bump from the interior but my burrs just kept getting clogged so much and it was so slow going that I abandoned it and redirected my efforts elsewhere.
Last edited by newbvetteguy; 01-26-2022 at 06:12 PM.
01-13-2022, 05:42 AM
