Hello!
I recently got a request on how to make your own intake manifold! Now this information/process is universal to any car but I'll be showing you how I made my own intake manifold for a B20A4!
Now for the owners of a B20A3/4, skip ahead to the end and just use my files, as they are designed for that.
Now the prerequisites for you to do this is to have a caliper, some sort of CAD program (I used Siemens NX12 but Fusion 360 is something i recommend as a NX license costs about 180-500 euro/dollars a month, I thank my school for sponsoring it) and some basic knowledge of how designing works and how to take measurements.
Now if you are here only for the measurements, then skip ahead to the ending of this post.
DO NOTE THAT THIS MANIFOLD CAN BE USED FOR ITB SETUPS.
Some ITB:s use the DCOE/DHLA flange to mount so you could buy some from i.e. FAJS and mount them to these flanges.
![Image]()
1. Measuring the intake to cyl head flange
First off you want to start with taking off your stock intake manifold, you can use this as a base for the form of the baseplate of the intake manifold to the cylinder head, and as a base for where the screw holes go. Now measure everything you can as a refrence, and you can sketch up a refrence on a piece of paper, or write down values directly on the intake manifold as i did.
![Image]()
Now the way I measured things, I used a base point on the intake manifold to then refrence other measurements further, for my case I started with the top bolt mounting point in the middle of the intake manifold, and then measured out from it. Essentially like this:
1. Top bolt hole diameter
2. Bolt hole edge to edge + 1 diameter = center to center from one bolt to another
3. bolt hole to inner port inner edge
4. port width
5. port edge to port edge
6. port width (every port should have the same size so kinda unnecessary.)
7. outer port outer edge to top outer bolt hole
And keep doing this on and on until you have just about all needed measurements, this depends on how advanced the geometry is and how confident you yourself feel about the measurements.
Personally I measured a million things and double- triple- quadruple checked everything, as I decided with going with CNC and you do not want such an expensive part to be wrong as no warranty will be available as it is your own design.
(Regarding my constricts method, here is a picture to understand better what I mean)
![Image]()
Now one key detail to notice when doing this yourself: MEASURE PORT DIAMETER FROM THE CYLINDER HEAD.
Reason is because some manufacturers make the port diameter of the intake manifold smaller than the cylinder heads port diameter, meaning less flow, the reason for this may vary, like easier tolerances during manufacturing or maybe better flow rates for the stock setup.
After taking measurements, let's sketch it in CAD.
2. Sketching in CAD
THIS IS NOT A GUIDE ON HOW TO SKETCH IN CAD, JUST SOME GENERAL ADVICE, GO LEARN THAT SOMEWHERE ELSE.
Now as a intake manifold is usually symetrical, you only have to sketch up half of it and then mirror it in the CAD software.
Some other advice I can give you is to keep it milimeters as that is what major CNC shops and CAD software use, at least in the rest of the world, dunno with the US.
Flange thickness is up to the weight of the rest of the system connected to these carbs, the flanges are load bearing, aka it supports the weight of the pipes and what is connected to them, so just don't make it 1-5mm, I used 10mm which some might argue is even too much, but better safe than sorry.
![Image]()
When it comes to the bolt holes of the intake manifold flange, of course you've gotta make the holes bigger than the bolt diameter itself, but don't make it too big, as that will result in a loose fit. The thread diameter of the intake manifold flange on the B20A4 is 8mm, so I made the holes 9mm, which gives it a snug fit.
![Image]()
On the stock manifold there is a gap between the two middle ports in the manifold to be able to create a constant vacuum pressure between all 4 ports, but we will be removing that gap so that air flow in the pipes are more efficient.
here is my process of making this flange:
![Image]()
Now I don't have any tools to measure the radius of the curves of this manifold so I just guesstimated them, they aren't big radii, and the bigger radii are on the outer side of the flange so they aren't too exact.
3. Measuring the ITB/Carb flange.
The sketches of the flanges for the DCOE/DHLA/whatever solex has can be found readily anywhere on the internet as long as you know what you are searching for.
Now my own theory of getting the right size of carbs/itbs, to get the right size diameter for the ITB:s/carbs, calculate the area of the manifold ports an round to the closest number of diameter of ITB:s/carbs.
Example: 37.5 height * 34 length = 1275 mm^2
square root of (1275 mm^2/pi)=20.14mm radius) = 40.29 mm diameter
There are a lot of other factors when choosing size but this is just an estimate, a good tool I used was a weber dcoe calculator, which gave me a lot of general specifications like what carb and what jetting and more. (Weber jetting calculator - en - USD)
Here is the DCOE/DHLA flange:
![Image]()
Once again the flange is 10mm thick.
4. Runners
This is the hardest part. Took me several days to figure out, but essentially all I did was use the spline tool to create the curvature i wanted in between the flanges, and connected one port to another with the Swept tool in NX, this might vary from software to software.
What I was left was a pair of thin runners which I then had to thicken, I made them the runners about 5mm thick for robustness, but thin enough that they wouldn't interfere with the bolts and one another.
![Image]()
![Image]()
The reason i had to make a spline for the Swept tool to follow was to get better flow as the runners being straight would be non optimal.
Of course this wouldn't be a problem if not the B20A was tilted at 18 degrees. As if the runners were straight, then the carbs would be pointing 18 degrees downward, and not function correctly as they are sidedraft carburetor.
5. Vacuum manifold outputs
On the bottom of the runners you want to put some ports to make a vacuum manifold for all the accessories that require vacuum i.e. brake booster and ignition advance. I just made some cylinders and united them to the manifold, then used the hole tool to make holes for 1/8 BSP, as this was the smallest diameter I could get from the local pneumatics store. If you don't do this, good luck braking and getting the engine to run properly.
![Image]()
6. Finalizing the design
As a final, optimize the design, chamfer edges, edge blend edges, make a draft with all port sizes and thread sizes (as the picture above) for the CNC manufacturer.
The most important parts is exporting a proper file, make sure to hide all components that are unnessecary, like sketches, the swept edge and more, just leave all the extruded parts and holes, then use the unite tool to unite the whole body so it becomes one single part.
7. Final notes
After that just export the part to the shops prefered file format, i.e. .stp, .f3d, .part etc.
Now regarding the material, I used Al6061, any other metals like steel and titanium (if you are crazy enough) work too, one just weighs more but is more reliable, and the other is better in every way except for the cost and heat exchanging properties. You could also use additive manufacturing (commonly known as 3d printing) to produce this part, but the quality of the surface finish might be worse (don't know, haven't tried it, just using plastic 3d printing as a refrence, but that is quite a primitive way of 3d printing at this point).
I used a surface finish of Ra 3.2 and 220 Grit after. The tolerances of the parts differ between designs, but I used DIN ISO 2768-M for my part (google for more details), as it is one of the industry standards.
ANY QUESTIONS ARE WELCOME, BE SURE TO ASK THEM IF YOU WANT TO!
I'll be adding the part files at a later date, one with more room between the left and right carb, one with another vacuum port threading and one with both. Currently the design is finalizing and I'm still optimizing it.
Edit: files are done and can be got by contacting me in dms.
TOTAL COST: ~600 USD for CNC and ~450 USD for 3D printed
I recently got a request on how to make your own intake manifold! Now this information/process is universal to any car but I'll be showing you how I made my own intake manifold for a B20A4!
Now for the owners of a B20A3/4, skip ahead to the end and just use my files, as they are designed for that.
Now the prerequisites for you to do this is to have a caliper, some sort of CAD program (I used Siemens NX12 but Fusion 360 is something i recommend as a NX license costs about 180-500 euro/dollars a month, I thank my school for sponsoring it) and some basic knowledge of how designing works and how to take measurements.
Now if you are here only for the measurements, then skip ahead to the ending of this post.
DO NOTE THAT THIS MANIFOLD CAN BE USED FOR ITB SETUPS.
Some ITB:s use the DCOE/DHLA flange to mount so you could buy some from i.e. FAJS and mount them to these flanges.
1. Measuring the intake to cyl head flange
First off you want to start with taking off your stock intake manifold, you can use this as a base for the form of the baseplate of the intake manifold to the cylinder head, and as a base for where the screw holes go. Now measure everything you can as a refrence, and you can sketch up a refrence on a piece of paper, or write down values directly on the intake manifold as i did.
Now the way I measured things, I used a base point on the intake manifold to then refrence other measurements further, for my case I started with the top bolt mounting point in the middle of the intake manifold, and then measured out from it. Essentially like this:
1. Top bolt hole diameter
2. Bolt hole edge to edge + 1 diameter = center to center from one bolt to another
3. bolt hole to inner port inner edge
4. port width
5. port edge to port edge
6. port width (every port should have the same size so kinda unnecessary.)
7. outer port outer edge to top outer bolt hole
And keep doing this on and on until you have just about all needed measurements, this depends on how advanced the geometry is and how confident you yourself feel about the measurements.
Personally I measured a million things and double- triple- quadruple checked everything, as I decided with going with CNC and you do not want such an expensive part to be wrong as no warranty will be available as it is your own design.
(Regarding my constricts method, here is a picture to understand better what I mean)
Now one key detail to notice when doing this yourself: MEASURE PORT DIAMETER FROM THE CYLINDER HEAD.
Reason is because some manufacturers make the port diameter of the intake manifold smaller than the cylinder heads port diameter, meaning less flow, the reason for this may vary, like easier tolerances during manufacturing or maybe better flow rates for the stock setup.
After taking measurements, let's sketch it in CAD.
2. Sketching in CAD
THIS IS NOT A GUIDE ON HOW TO SKETCH IN CAD, JUST SOME GENERAL ADVICE, GO LEARN THAT SOMEWHERE ELSE.
Now as a intake manifold is usually symetrical, you only have to sketch up half of it and then mirror it in the CAD software.
Some other advice I can give you is to keep it milimeters as that is what major CNC shops and CAD software use, at least in the rest of the world, dunno with the US.
Flange thickness is up to the weight of the rest of the system connected to these carbs, the flanges are load bearing, aka it supports the weight of the pipes and what is connected to them, so just don't make it 1-5mm, I used 10mm which some might argue is even too much, but better safe than sorry.
When it comes to the bolt holes of the intake manifold flange, of course you've gotta make the holes bigger than the bolt diameter itself, but don't make it too big, as that will result in a loose fit. The thread diameter of the intake manifold flange on the B20A4 is 8mm, so I made the holes 9mm, which gives it a snug fit.
On the stock manifold there is a gap between the two middle ports in the manifold to be able to create a constant vacuum pressure between all 4 ports, but we will be removing that gap so that air flow in the pipes are more efficient.
here is my process of making this flange:
Now I don't have any tools to measure the radius of the curves of this manifold so I just guesstimated them, they aren't big radii, and the bigger radii are on the outer side of the flange so they aren't too exact.
3. Measuring the ITB/Carb flange.
The sketches of the flanges for the DCOE/DHLA/whatever solex has can be found readily anywhere on the internet as long as you know what you are searching for.
Now my own theory of getting the right size of carbs/itbs, to get the right size diameter for the ITB:s/carbs, calculate the area of the manifold ports an round to the closest number of diameter of ITB:s/carbs.
Example: 37.5 height * 34 length = 1275 mm^2
square root of (1275 mm^2/pi)=20.14mm radius) = 40.29 mm diameter
There are a lot of other factors when choosing size but this is just an estimate, a good tool I used was a weber dcoe calculator, which gave me a lot of general specifications like what carb and what jetting and more. (Weber jetting calculator - en - USD)
Here is the DCOE/DHLA flange:
Once again the flange is 10mm thick.
4. Runners
This is the hardest part. Took me several days to figure out, but essentially all I did was use the spline tool to create the curvature i wanted in between the flanges, and connected one port to another with the Swept tool in NX, this might vary from software to software.
What I was left was a pair of thin runners which I then had to thicken, I made them the runners about 5mm thick for robustness, but thin enough that they wouldn't interfere with the bolts and one another.
The reason i had to make a spline for the Swept tool to follow was to get better flow as the runners being straight would be non optimal.
Of course this wouldn't be a problem if not the B20A was tilted at 18 degrees. As if the runners were straight, then the carbs would be pointing 18 degrees downward, and not function correctly as they are sidedraft carburetor.
5. Vacuum manifold outputs
On the bottom of the runners you want to put some ports to make a vacuum manifold for all the accessories that require vacuum i.e. brake booster and ignition advance. I just made some cylinders and united them to the manifold, then used the hole tool to make holes for 1/8 BSP, as this was the smallest diameter I could get from the local pneumatics store. If you don't do this, good luck braking and getting the engine to run properly.
6. Finalizing the design
As a final, optimize the design, chamfer edges, edge blend edges, make a draft with all port sizes and thread sizes (as the picture above) for the CNC manufacturer.
The most important parts is exporting a proper file, make sure to hide all components that are unnessecary, like sketches, the swept edge and more, just leave all the extruded parts and holes, then use the unite tool to unite the whole body so it becomes one single part.
7. Final notes
After that just export the part to the shops prefered file format, i.e. .stp, .f3d, .part etc.
Now regarding the material, I used Al6061, any other metals like steel and titanium (if you are crazy enough) work too, one just weighs more but is more reliable, and the other is better in every way except for the cost and heat exchanging properties. You could also use additive manufacturing (commonly known as 3d printing) to produce this part, but the quality of the surface finish might be worse (don't know, haven't tried it, just using plastic 3d printing as a refrence, but that is quite a primitive way of 3d printing at this point).
I used a surface finish of Ra 3.2 and 220 Grit after. The tolerances of the parts differ between designs, but I used DIN ISO 2768-M for my part (google for more details), as it is one of the industry standards.
ANY QUESTIONS ARE WELCOME, BE SURE TO ASK THEM IF YOU WANT TO!
I'll be adding the part files at a later date, one with more room between the left and right carb, one with another vacuum port threading and one with both. Currently the design is finalizing and I'm still optimizing it.
Edit: files are done and can be got by contacting me in dms.
TOTAL COST: ~600 USD for CNC and ~450 USD for 3D printed
