J. Phil Thien's Projects

Absolute collection of fines for my system seems to be the function of the filter.  I crudely tested my system with flour and collected ~2/3 in the baffle can.  I have to use pressurized air to blow into the filter from the outside ~1 every two years to remove fines.  My recollection of the benefits of the thien baffle was that it collected the bigger stuff with a smaller pressure drop and smaller than the cyclone but wasn't as good as collecting fines as a cyclone. The only time big stuff gets past is when my can fills up and I don't notice it. Your idea of tapering the outside is interesting.  Since your top hat is taller, the inlet to the internal pipe to the blower could be set at a higher height so that the distance from the pipe to the tapered wall is similar to the smaller design typically shown here.  However, this is just a guess.

Your scaling calculations appear correct for the assumptions you made, however other considerations may be needed.  Does the diameter need to change of the top hat? I haven't seen any info on 8" piping so you may need to experiment. The distance from the 8" pipe to the tangential inlet is narrower and could affect dust collection. Also, an 8" trunk line for a vacuum system requires a large amount of HP to keep the air moving at 45 mph to keep saw dust airborne.  Too slow and dust will collect in the 8" line. The advertised CFM on dust collectors is with no load and you can't trust them for your system design.  Scaling my HF 5" existing piping system would mean that I would need a 4-5 HP motor with an 8" line to keep dust from accumulating in the trunk line. The velocity of the air at the 4" intakes would be much higher than needed. The pressure losses of the diameter, length, number of elbows, and number of lines affect the HP requirements as well.  The top hat design is made for small DIYers with small systems.  The size of this system appears to be more of a commercial system.  Some DIYers with big shops will use a few portable smaller systems with a top hat rather than this large system with a lot of piping.  If you haven't already, look at all your needs and what designs can meet them before you go much farther. Also, small affordable cyclones are now available which were not available when Phil proposed this design.  If you need more information, there is a huge amount of info on the internet on any of these subjects with simple searches. 

Response to BigSteve
I modified my HF with the shaft vertical several years ago and it runs fine.

Your last picture is looking in the inlet.  Two possible improvements.
- The bellmouth pipe looks like it is almost to the bottom surface.  If it is, the bottom of this pipe should be about half height or higher.
- The internal ceiling appears to have an opening to allow the bellmouth to be inserted which leaves a big gap between the pipe and the top plate. Airflow may be better if it was plugged or filled.  In addition, this connection needs to be air tight as others had mentioned for all connections.

The air flow that I have is sufficient to lift the chips.  I added the Rikon because I added a downdraft table near the DC.  For a Down draft table you need high CFM and this was an easy mod to get higher CFM.

Sorry the system send this before I was done.  I hit a tab for spacing but it went to post.
Try a Wynn filter for the top HF bag. Collects finer dust and has more flow area.
Look for possible leaks and tape them up.

I did the same setup about 5 yrs ago.  These are the things I did based on internet info and a flow meter on my system.
Need a blast gate at each tool.
Used metal hvac pipe.
All routing off the main 5" line should use Ys not T intersections.
They now have plastic 5" Ys reasonably priced.  I made mine from two straight metal pipes pipes.
Found 5" pipe at local wholesale distributor cheaper and better than box store
Use 4" drop downs at all tools.  maximizes velocity to counter gravity.  Actually your plan of 5" drop downs at the end of the run may have problems.  It may work close to the collector.
Minimize 90 deg turns
Used two adjustable 90 deg elbows set at ~45 to make a more gradual 90 deg turn
Tape all joints.  This added roughly 15 mph at the end of the run. For adjustable metal elbows tape each slip joint as well.
Minimize flexible pipe.
If the flow is weak at the tools on the end, you may want to try these:

change the HF impeller.  See internet for Rikon replacement.

You may find the size you need by looking up register boxes.  I got mine custom made (2015) from spiral Manufacturing but I could not find the page today to make a new order from. My collector is described in "HF DC spiral collector with flow redirection" and I had a 5" round transition to a 2"x10" rectangle.  The cost for a wide mouth and the transition with shipping was ~$90.  You may be able to find a local sheet metal shop to do it.

see this post
HF DC spiral collector with flow redirection

Unless you have to pay for shipping HVAC 5" metal pipe is a reasonable cost.  You may have to look at a local hvac commercial distributor rather than the box stores.  Plastic 5" wyes are now available from woodcraft for ~$15.

Too much opening leads to too much flow in the lower collector.  I used a spiral design and only had about 30degrees closed. I was getting too much flow in the lower can.  The dust in the lower can was all in the center rather than the outside.  I closed it back to about 120 and it seems better.  I agree with R2 stay with the 120.  The turbulent entrance expands into the upper cavity before it has a chance to expand into the lower can.

quick thoughts:
The diesel exhaust temperature (>500F from google) could burn up or melt the typical thien baffle materials.  So make it out of non flammable materials suitable for your exhaust temperature.
The soot may be sticky and collect on the perimeter of the collector, so be prepared to clean it a lot or make it out of something soot doesn't stick to.  The added friction may also degrade the collection performance with time.
The back pressure of the baffle may increase the exhaust pressure and reduce the efficiency of the diesel.
Unique application, be careful.  Who knows, what you develop could be the next device we see on trucks.
If it does stick to a common, cheaply available product, make the air spin within a cylinder of the stuff and pass the soot through it, and throw it away when full like a filter but air flow is open.

Comments based on my build:
1) If the baffle is before the impeller, you will need something inside the bag to keep the bag in the collector can from collapsing.  Wire outside the bag would have to have a connection to the bag.   I tried a plastic bag without support and it sucked up to the baffle slot. I do not use a bag.  The mesh inside the bag makes the bag non disposable so you will have to pour it out anyway, so why use a bag?
2) If your run is more than 10 feet, use 5" hvac pipe.  Its relatively inexpensive. Wye connections are expensive but are not too hard to make.  Make sure all the seams are taped.  If you don't, you will lose all the benefit of pipe.  Reduce to 4" and run a few feet down to your equipment.
3) As you noted side entry is best.
4)  I routered a circle groove the same diameter as my can into the bottom of the top hat and I put spongy one sided sticky insulation into the groove.  I wedge the bottom of can up to seal it tight.  If it leaks you will lose air flow.

Some thoughts to ponder:
- Larger pipe is so much better for high flow that you may look at other alternatives like:
--Figure out where the ceiling joists run.  Situate your DC so that you can remove the ceiling surface until you see the joists and run the pipe between the joists.  If you care, you can redo the ceiling and only have the pipe entrances and exits showing.  I only had to remove insulation for mine.
--Any PVC bigger than 4" gets really expensive for the connectors.  I used 5" HVAC ducts, used two 90 deg elbows to make more gradual turns. Tape all joints; it makes a huge difference in air flow.
--Y connectors are expensive and/or not available but you can make them out of straight metal duct.  Online calculators can give you the patterns to cut. It doesn't have to be exact, there is tape and putty to use to make it air tight. 
-The only negative of a 10" outlet that I can think of is that with these high flows it may be too close to the inlet flow near the perimeter (only 6.5" away) and dust may bypass the collector.  This may be one of those things that you try it and if it doesn't work well enough, try something else like a tapered pipe that's 10" at the top and maybe 6-8" at the bottom.
-You have so much flow that maybe a cyclone may be better.
-A large trunk line to the collector with 4" feed lines off it would solve the dual 4" inlet pipe issue.
-It sounds like you will have enough flow to have more than one 4" line active at once which can be effective for tools other than a table saw like a sanding table which needs very high flow.
-If you are scientifically inclined, there is information to estimate the effect of using different size pipe runs.  In the end, its a comparative calculation and you will have to guess the effect unless you know the CFM characteristics of your DC versus pressure drop.
Everyone's situation is different with space, layout, tools, $, time, DC, etc.  I hope some of these ideas may be useful for your situation.

Oops.  I meant to add the advertised picture at the following location. click on green picture.
http://www.bridgecitytools.com/default/tools/dust-collection/gyro-air.html
or look at attachment.