I had to laugh when this came up on my suggested videos, because a week or two back I was picking apart some post from some company showing off fancy-assed CFD of their manifold - flowing out all ports at the same time. Which, if that's happening in the engine, bad stuff is going on, the way I figure. So kudos, and subscribed to see where this goes.
@@half_tilt_boogie How would you actually try to simulate all the things happening in the magnifold within two full rotations of the camshaft? Or what would be the steps to get to a more and more realistic result? For example, does it even make sense to try to simulate the transition from one runner to another, or valves opening and closing, without also considering the pistons moving up and down?
Cool video. Like you said, intake and exhaust manifolds are a very chaotic system once scavenging / resonance / pressure waves etc are accounted for. Makes me wonder what kind of modelling and analysis tools the OEM's and race teams get to play with? I imagine the formula 1 manufacturers have pretty robust modelling software, rather than just throwing millions of dollars in prototype parts at the car and seeing what sticks
Great video, thank you! Really surprised to see that raised velocity stacks performed worse than integrated ones. I always thought that rasing stacks is important to improve flow because of issues with boundary layer.
@@half_tilt_boogie Thats exactly I was thinking about at the moment :) Also making the inlet wher throttle body angled to the bottom surface where cyl. inlets are should improve it as well.
Taper the plenum going from 1st to 4th cylinders. Reduced volume = more speed. This will let you compensate for the 3% loss. May be more difficult to fabricate though, since it will not be a simple cylinder.
Good video, decent design. Do you have any design suggestion for ensuring the rearmost ports get a more equal flow? A tapered design for the volume/plenum is the usual solution. I've been doing a lot of reading and research on inlet manifold design, Helmholtz resonance and David Vizard's substantial work in this field. I understand a 7º taper and bellmouth radius of 1/4 of the port entry diameter should yield a gain in flow of a few percent. The port entry radius is also better slightly flattened... or the surface around (and dividing) each port entry more semi-eliptical rather than semi-circular. Could you repeat this with above recommendations to compare the improvement?
Instant subscription! This is a topic close to my heart and hearing someone say resonance is a factor is great! I have got a question though How far has cfd developed? Have you seen or has anyone out there got the ability to simulate cam profile and resonance along with the firing order as too often hear people talking about cfm of intakes or heads,with little or no reference to resonance and the fact it’s filling a cylinder with a moving piston ,having seen higher hp from lower flowing setups I think that 4 stroke resonance although not as important as two stoke should be considered all the same,no doubt the top end of developers would like f1,
@@half_tilt_boogie couldn’t agree more,I’m a mechanic and do a fair bit of tinkering and tune and modify a wide variety of engines and myself being basically computer handicapped and generally not well educated find this cfd stuff fascinating! But consider hands on the only way, I’ve found in real world testing that intake resonance benefits can be easily seen with an engine running under load and shaping/modelling the intake simply by introducing an object like a concave dish or even flat plate in front of the bell mouths improvement of flow at particular rpm can be improved hugely, having said that usually I just turbo it!
Tom! You are honestly the only person I've been able to locate that knows how to operate Flowsquare+ to a high level. I have a complete understand of setting up models, but I do have some questions for you regarding your simulations: 1. Did you define the inlet and outlet by aligning the .bmp to the STL you imported? Seems to be the only way for me to get this to work. 2. Are the animations from Paraview? Any tips on that? That portion of this entire process seems very unrefined. 3. Lastly; am I not mistaken in that exports are defined every N # of time steps? It appears your animations are every 50, yes? That is controlled via param.txt, right? I am admittedly confused with how to export data from this software. It seems very powerful after getting multiple simulations to run.
Another thing is I've found on bench testing the throttle plate should be modeled (being open and in the air stream) and the inlet pipe going into it can have a significant difference on manifold distribution. ( Length and angle of approach to TB etc)
@@half_tilt_boogie Yes just in depression and pressure - I didn't have anything that verified flow, but it was significant enough through standard observation. I'm sure each manifold responds differently, but on this intake manifold it had an effect on all 6 ports. What it appeared to show was air not having enough time to make turns. Having slight changes to piping going into the TB biased the fluid stream. It didn't, say, fix the unequal distribution changing the position. Rather either exaggerated the problem or lessened it. Think of it like an equalizer where you can only make so many dB changes. You couldn't zero out all of the imbalances (at least not in this manifold) but you could improve it. I have no idea if any power change would result.
This is the another level of tuning. Am searching for the answer for my question and finally got it. With this same setup na vs forced induction comparison. Can you do it?
but if the engine is running it wouldnt have atmospheric pressure in the plenum. that would be on the opposite side of the tb. in the plenum it would have a partial vaccuum if engine is running. dont know how much itll affect results , but it might
Pretty good video and explanation!, I was looking for alternatives to simulate an intake manifold, and your previus video came as the first result. Thanks for introducing me to Flowsware! it lokks pretty neat. I Just Suscribed. :)
My gene 3 stock Subaru manifold is beautiful, aerodynamic, very small plenum that's completely smooth. My gen 4 Subaru has a black ABS plastic with smaller diameter runners, a bigger plenum that has serious protruding plastic squares that I would think cause serious drag/turbulence. Yet the newer gen makes more HP/torque. I'm befuddled.🤯
Can you do one for motorcycle plenum? For example for hayabusa, because there is not too much space, so we can’t do long runners, so we never know if we need a big plenum or smaller and big runners
Great video! Would arranging the runner inlets in the plenum in firing order improve or hinder the flow? For example, consider arranging a 4 cylinder plenum runners in the 1-3-4-2 order from front to back.
@@half_tilt_boogie I think the interesting part is where the air goes from feeding runner 1 to feeding runner 3, as opposed to going from 3 to 4. I think there is potential benefit in making the travel time/distance between runners the same. The dynamic cfds show this issue much more clearly than the static plots.
Tune the length of intake port runners from the throttle opening. As you would place the collector on a header. Throttle plate would have a balanced signal if it was equal distance from each intake port on other words... Ford and other manufacturers tried the dual port w valve vanes on the V6 it's a challenge on an in line engine for sure. GM twin cam from the Late 90', for example 97'Sunbirds 2.4 had an equal runner small plenum intake manifold... It looked like up side down header pipe, but on the intake side of the head. That engine my opinion suffered small chains running the balance shafts.
Thank's for this very informative video! It really explains a lot. Is it possible that you would share your model and configuration files for this simulation?
Nice content. Should run a boosted (over atmospheric) run against this one. There are so many schools of thought out there. Modern trends abandon resonance for 'shortest path' (response) and overall flow (design). Some say plenum volumes makes no difference under boost. No resonance tuning benefit (just add more boost).
@@half_tilt_boogie Yes, I agree, atmospheric IS boost already I keep trying to tell people that. But I have observed that greater pressure (density) tends to delete the effects of ports or cams being too large (at xxxx rpm NA) vs boosted. Other times not at all, or if so it had to be over a certain threshold. Sorta like when people run Vtec on, let's say an F20C, turning it on much sooner than stock might result in less power, especially if you're trying to switch the profile thousands of rpms earlier (while NA) But of course under boost, if the turbo has already made appreciable boost you can and should switch to the larger lobes thousands of RPM sooner)
holy shit how can this video only has 1k view? it deserves way more. Love your video man !
Your video from two years ago was recommended to me. Checked for the updated version and saw it was done yesterday lol
@@half_tilt_boogie love watching videos like this, keep it up
I had to laugh when this came up on my suggested videos, because a week or two back I was picking apart some post from some company showing off fancy-assed CFD of their manifold - flowing out all ports at the same time. Which, if that's happening in the engine, bad stuff is going on, the way I figure. So kudos, and subscribed to see where this goes.
@@half_tilt_boogie How would you actually try to simulate all the things happening in the magnifold within two full rotations of the camshaft? Or what would be the steps to get to a more and more realistic result? For example, does it even make sense to try to simulate the transition from one runner to another, or valves opening and closing, without also considering the pistons moving up and down?
Cool video. Like you said, intake and exhaust manifolds are a very chaotic system once scavenging / resonance / pressure waves etc are accounted for.
Makes me wonder what kind of modelling and analysis tools the OEM's and race teams get to play with?
I imagine the formula 1 manufacturers have pretty robust modelling software, rather than just throwing millions of dollars in prototype parts at the car and seeing what sticks
Great video, thank you! Really surprised to see that raised velocity stacks performed worse than integrated ones. I always thought that rasing stacks is important to improve flow because of issues with boundary layer.
@@half_tilt_boogie Thats exactly I was thinking about at the moment :) Also making the inlet wher throttle body angled to the bottom surface where cyl. inlets are should improve it as well.
Taper the plenum going from 1st to 4th cylinders. Reduced volume = more speed. This will let you compensate for the 3% loss. May be more difficult to fabricate though, since it will not be a simple cylinder.
Little bit more fabrication but definitely not impossible. Just cut the tube in half and add some sheet metal in between with right shape.
Good video, decent design.
Do you have any design suggestion for ensuring the rearmost ports get a more equal flow? A tapered design for the volume/plenum is the usual solution.
I've been doing a lot of reading and research on inlet manifold design, Helmholtz resonance and David Vizard's substantial work in this field. I understand a 7º taper and bellmouth radius of 1/4 of the port entry diameter should yield a gain in flow of a few percent. The port entry radius is also better slightly flattened... or the surface around (and dividing) each port entry more semi-eliptical rather than semi-circular.
Could you repeat this with above recommendations to compare the improvement?
Very good work and explanation.
Instant subscription! This is a topic close to my heart and hearing someone say resonance is a factor is great! I have got a question though How far has cfd developed? Have you seen or has anyone out there got the ability to simulate cam profile and resonance along with the firing order as too often hear people talking about cfm of intakes or heads,with little or no reference to resonance and the fact it’s filling a cylinder with a moving piston ,having seen higher hp from lower flowing setups I think that 4 stroke resonance although not as important as two stoke should be considered all the same,no doubt the top end of developers would like f1,
@@half_tilt_boogie couldn’t agree more,I’m a mechanic and do a fair bit of tinkering and tune and modify a wide variety of engines and myself being basically computer handicapped and generally not well educated find this cfd stuff fascinating! But consider hands on the only way, I’ve found in real world testing that intake resonance benefits can be easily seen with an engine running under load and shaping/modelling the intake simply by introducing an object like a concave dish or even flat plate in front of the bell mouths improvement of flow at particular rpm can be improved hugely, having said that usually I just turbo it!
Tom! You are honestly the only person I've been able to locate that knows how to operate Flowsquare+ to a high level. I have a complete understand of setting up models, but I do have some questions for you regarding your simulations:
1. Did you define the inlet and outlet by aligning the .bmp to the STL you imported? Seems to be the only way for me to get this to work.
2. Are the animations from Paraview? Any tips on that? That portion of this entire process seems very unrefined.
3. Lastly; am I not mistaken in that exports are defined every N # of time steps? It appears your animations are every 50, yes? That is controlled via param.txt, right? I am admittedly confused with how to export data from this software. It seems very powerful after getting multiple simulations to run.
Another thing is I've found on bench testing the throttle plate should be modeled (being open and in the air stream) and the inlet pipe going into it can have a significant difference on manifold distribution. ( Length and angle of approach to TB etc)
@@half_tilt_boogie Yes just in depression and pressure - I didn't have anything that verified flow, but it was significant enough through standard observation. I'm sure each manifold responds differently, but on this intake manifold it had an effect on all 6 ports.
What it appeared to show was air not having enough time to make turns. Having slight changes to piping going into the TB biased the fluid stream.
It didn't, say, fix the unequal distribution changing the position. Rather either exaggerated the problem or lessened it.
Think of it like an equalizer where you can only make so many dB changes. You couldn't zero out all of the imbalances (at least not in this manifold) but you could improve it.
I have no idea if any power change would result.
This is the another level of tuning. Am searching for the answer for my question and finally got it.
With this same setup na vs forced induction comparison. Can you do it?
Awesome that fella! Impressive stuff 🙌✅👍
but if the engine is running it wouldnt have atmospheric pressure in the plenum. that would be on the opposite side of the tb. in the plenum it would have a partial vaccuum if engine is running. dont know how much itll affect results , but it might
Pretty good video and explanation!, I was looking for alternatives to simulate an intake manifold, and your previus video came as the first result.
Thanks for introducing me to Flowsware! it lokks pretty neat. I Just Suscribed. :)
My gene 3 stock Subaru manifold is beautiful, aerodynamic, very small plenum that's completely smooth. My gen 4 Subaru has a black ABS plastic with smaller diameter runners, a bigger plenum that has serious protruding plastic squares that I would think cause serious drag/turbulence. Yet the newer gen makes more HP/torque. I'm befuddled.🤯
Can you do one for motorcycle plenum? For example for hayabusa, because there is not too much space, so we can’t do long runners, so we never know if we need a big plenum or smaller and big runners
Can you do a video for a centre feed style plenum?
Great video, thanks
Any interest checking a dual plenum?
I don't understand what he's talking about but it sounds very cool
Edit : subbed because the quality is 11/10
Great video! Would arranging the runner inlets in the plenum in firing order improve or hinder the flow? For example, consider arranging a 4 cylinder plenum runners in the 1-3-4-2 order from front to back.
@@half_tilt_boogie I think the interesting part is where the air goes from feeding runner 1 to feeding runner 3, as opposed to going from 3 to 4. I think there is potential benefit in making the travel time/distance between runners the same. The dynamic cfds show this issue much more clearly than the static plots.
@@CheekyNinja Without variable runner length it would help in some RPM but be contraproductive in other RPM...
@@vasek987 Evening out the mass distribution between the cylinders should help at all rpm. Runner length is a separate topic :)
@@CheekyNinja Yup, I was thinking about something completely different. My misunderstanding ;)
Tune the length of intake port runners from the throttle opening. As you would place the collector on a header. Throttle plate would have a balanced signal if it was equal distance from each intake port on other words... Ford and other manufacturers tried the dual port w valve vanes on the V6 it's a challenge on an in line engine for sure. GM twin cam from the Late 90', for example 97'Sunbirds 2.4 had an equal runner small plenum intake manifold... It looked like up side down header pipe, but on the intake side of the head. That engine my opinion suffered small chains running the balance shafts.
Thank's for this very informative video! It really explains a lot. Is it possible that you would share your model and configuration files for this simulation?
@@half_tilt_boogie Do you have an email or any other way to send you a PM?
I would really appreciate it :)
Would tapered intake runners make a difference? Thx!
Nice content.
Should run a boosted (over atmospheric) run against this one.
There are so many schools of thought out there.
Modern trends abandon resonance for 'shortest path' (response) and overall flow (design).
Some say plenum volumes makes no difference under boost. No resonance tuning benefit (just add more boost).
@@half_tilt_boogie Yes, I agree, atmospheric IS boost already I keep trying to tell people that.
But I have observed that greater pressure (density) tends to delete the effects of ports or cams being too large (at xxxx rpm NA) vs boosted. Other times not at all, or if so it had to be over a certain threshold.
Sorta like when people run Vtec on, let's say an F20C, turning it on much sooner than stock might result in less power, especially if you're trying to switch the profile thousands of rpms earlier (while NA)
But of course under boost, if the turbo has already made appreciable boost you can and should switch to the larger lobes thousands of RPM sooner)
Could you share the dimensions?
The yt algorithm is high again, but good video :D
Open have and create your cambered ..turbo muffler weld it back dont peoplez or the government
It's completely wrong of course because air flows in waves.
k, i guess you did not even watch the video