Se have slow flow.... so what makes the rocket, uh, rockety?

We put together a test system using firebrick, 8" I.D. round insulated heat riser, 55 gal steel drum, 8" exhaust pipe (16' w/ three 90's), following Ianto's book pretty closely. We have slow flow through the system with some worrisome backsmoke.

We want to understand the dynamics better. What makes the rocket force? How can we tweak the system to get a stronger flow?

The only deviations we made from a traditional design is using the barrel right side up with a removable lid, and cutting holes in the bottom for the heat riser and 8" outflow directly connected to the bottom of the barrel. The heat riser has about 1.5 inches of rock wool blanket insulation. We dry stacked the blocks and sealed joints with clay/sand mortar. The burn tunnel is about 6" high x about 7.25" wide. It's not extra long either.

Don't get me wrong, we DID get a positive flow through the system, but it was weak, not as rockety as I would want to pull more fire and smoke in that direction. It was too easy for fire and smoke to go vertically up the feed area. And we still have more exit flue to add to go up and out a wall in the building.

So, what makes the rocket power?

Photo: http://www.flickr.com/photos/9867280@N07/8178815692/in/photostream

What is your gap between the top of the heat riser and the barrel? Ianto's book explains the math that an 8" system needs a minimum of 2" and that more than 3" would slow the draw.

3.5 inches. Would a half inch make a difference? I still don't understand the the dynamics. Why is the spacing of that gap so critical??

First I must recommend Ianto Evan's book here http://www.rocketstoves.com I bought the download and printed it out so i could use it in reference, get it dirty n not worry.
The space between the riser and barrel is referred to as the "imaginary ring". Picture the inner riser pipe diameter extending up to the barrel, that "ring" must be measured and CSA (cross sectional area) calculated. An 8" pipe has a circumference of 25.12" (Pi x Diameter) and area of 50.24 sq" (Radius squared x Pi). That area (50") must be maintained throughout. At the imaginary ring, it it calculated as the circumference (25.12) X height (3.5) = 87.92 sq". The larger the area, the slower the flow, thus less rockety. This is good in some parts of the RMH but bad in others. This location has a tight tolerance to maintain suction of gasses and preventing smoke back.
Hope that helps