CFD/CAD/Thread

[SpeierRacingHeads]

This is a perfect learning example of how upstream changes affect downstream event. The reverse can occur too of course.

You've achieved a better velocity gradient which is a good thing and will likely produce a quieter port but as you pointed out the vorticity has increase in the bowl. I think you'd both agree that the area right before the throat/seat on the short side is the biggest concern.

Does your program take into account back angles on the valve? If so, can you change the width on them? Turbulence is also occurring around the valve stem. It would be interesting to see if widening and/or raising the teardrop at different points around the guide has a positive effect.

How much area did you lose with this change? Did you increase area elsewhere to compensate?

I would like to see what occurs at different lift points especially where peak CFM demand occurs and higher.

Speaking of port sound. This is the port. Notice the sound. Also notice when I stick some type of intruder how it reattaches to the flow.

https://www.youtube.com/watch?v=EmK57ArxG7g

[SchmidtMotorWorks]

What would filling the roof approximately like this achieve?

fillroof.jpg

Is the orange filling or removing?

[JC565Ford]

It might like the floor to have a bit of a dip before the crest of the short side. Sometimes the line of sight on the floor is not the best thing.

Can you use an image editor to draw some curves to explain what you mean?

I might be able to do that later, but for now look at the port labeled "old" look at the curve in the floor under the word "old" in the picture posted. That's the area I'm talking about.

Might give you guys some ideas:
13:00 - end
https://www.youtube.com/watch?v=qhsTQn0uUOQ&t=900s

Also - see 3:55 - 4:30 and 17:00 - 20:30
https://www.youtube.com/watch?v=lcxp9ufC1DY

[SpeierRacingHeads]

Thought I'd show you also. I did rework a port floor and had it rescanned.

[JC565Ford]

Thought I'd show you also. I did rework a port floor and had it rescanned.

blue is laid back, and did it flow better that way ?

there is a relationship between the floor and the roof line at the turn. if the relationship stays in balance you can keep things under control.

[BLSTIC]

What would filling the roof approximately like this achieve?

fillroof.jpg

Is the orange filling or removing?

Orange is fill

I figured they were two areas with a relatively sharp turn, and the one after the guide was showing as low velocity as well as the wall pointing the air towards the back face of the valve rather than out the curtain.

I need to look at the port properly again from all the different angles...

*edit: yup I definitely need to download these and view them in 3d. Having trouble visualising flow through the guide "scallops". I'm thinking they need to change somehow but I haven't figured out how yet.

[SchmidtMotorWorks]

*edit: yup I definitely need to download these and view them in 3d. Having trouble visualising flow through the guide "scallops". I'm thinking they need to change somehow but I haven't figured out how yet.

Yes, anyone that has not done that, should do so, it is worth the minimal effort.

[F120]

How high can you make the intake port?

I haven't been able to make a port yet that flowed closer than 2% of the flow bench.

Machining from CAD surfaces does make it easier.

[SpeierRacingHeads]

Thought I'd show you also. I did rework a port floor and had it rescanned.

blue is laid back, and did it flow better that way ?

there is a relationship between the floor and the roof line at the turn. if the relationship stays in balance you can keep things under control.

yes blue flowed better. I keep all the shrinkage of the port behind the apex and down into the throat. The back wing is actually shrinking the throat.

[SpeierRacingHeads]

How high can you make the intake port?

I haven't been able to make a port yet that flowed closer than 2% of the flow bench.

Machining from CAD surfaces does make it easier.

I'm at the height IMO. Also, no weld in the rocker area.

[SchmidtMotorWorks]

To save time I went ahead and made some changes.

The changes that made the most difference are:
1. Reshaping the pinch area, both sides, top and floor.
2. Reshape the manifold to make the top and floor tangent to the port.
Increase the radius of the turn at the end
3. Reshape the long side of the turn to reduce volume to conform to volume rule.
4. Reshape the guide boss to fill volume and reduce drag

This reduced the cylinder vacuum considerably.
The challenge is that this requires greater pushrod offset, probably about 0.250"
So we have to make a decision of whether to use a rocker with more offset or try to make a lower floor substitute.
Doubtful that the lower floor will work as well.

To make this design work, a slightly different design is needed for left/right inner/outer are required (4 different ports).
A manifold with perpendicular runners would make that unnecessary and probably improve performance.

JT files in next post.

Side view

side.png

Top view

top_pushrod soften turn.png

Reflections show smoothness

reflection 1.png

reflection 2.png

Velocity section

velocity_section_long.png

Vorticity

vorticity_section_long.png

Pressure Vacuum reduced from 5.297psi to 4.832psi

pressure.png

[SchmidtMotorWorks]

Here are some 3D JT files for the images above
The JT with arrows is too large to upload.
If anyone is very interested in that, I can make a shared Google drive (requires that you have a Google account)

streamlines.zip

[SpeierRacingHeads]

Can you put a size to how much bigger the pushrod area is?

[SchmidtMotorWorks]

Can you put a size to how much bigger the pushrod area is?

Sure, I was already planning to make a graph of the areas.

[SchmidtMotorWorks]

Here is something fascinating that I found today.
It looks like CFD will advance more in the next year than it has in the last 50.

https://www.youtube.com/watch?v=2Bw5f4vYL98