Having thought about this thread over the weekend - I admire the laboratory approach that you guys are taking, if mechanically assisted flue systems are not explored, then a definitive conclusion based on experimentation will be lacking in the development of the RMH. However, I cannot help wondering why just designing a flue that is efficient in the first place is not a prerequisite for any heating system that burns 'stuff'. The principle objective of a flue is twofold; firstly to take away the poisonous fumes (CO, CO2 etc.) and secondly to provide a draw of fresh air containing that useful ingredient for combustion O2.
Is the problem that the RMH is too effective at drawing heat our of the exhaust gases that by the time the exhaust flue in the designed system starts the 'exit' path that they are too cold to create a thermal draw to the outside?
Clearly this state would exist upon start-up and if not designed into the system would mean that the RMH would never be easy to light. The traditional solution to this problem would be to have a bypass baffle in the flue circuit that would enable the initial flue gasses to bypass any heat sink route and exit directly to the 'chimney' part of the system. Once started the baffle would be set in the usual position (perhaps gradually) to enable the whole flue to be used. I have seen examples in country houses in England where in the 18th century, to maintain the architectural integrity, a house would be built without chimneys in the main building, but a separate chimney being built higher up on a nearby hill (Chatsworth House is one of these examples), of course, the initial draw of the cold system meant that no fires could be started in the house without pre-heating the flue, so the solution was to have miniature fires in 2 or 3 places in the flue connecting the house with the stack - once warm, the draw in the house was sufficient to maintain the overall system of fires.
So another solution could be a smaller 'fire hole' in the flue that enables the initial draw.....
I'm just offering these alternative ideas because any system that requires additional technology to keep it running relies on expensive and energy-intensive production processes, for me the simpler ideas would take preference over the more complicated ones due to their reliance on less resources. I love the idea that the RMH could produce electricity to charge batteries etc., but is the problem we are looking for here simply that we are designing RMH flue circuits for steady-state operation which has created a problem with the initial start-up do to the lack of draw in the flue? If it is a question of aiding flue efficiency the surely a passive improvement is more durable (and cheaper) than an active 'fan-assisted' type design?
By the way, I installed an Okofen wood pellet boiler last year and have just realised that the initial circuit within the burner/water boiler section is similar to the RMH. Initial burn with controlled air feed produces the primary combustion, with an additional air injection a little way above the base 'pyramid' for secondary combustion. the hot gasses then pass down the internal surface of the boiler and then up through holes within the water jacket, then out of the flue (see http://outilssolaires.com/pro/fabricants-distributeurs/okofen-france+e1051.html
). The temperature of the outgoing gasses is rarely above 80°C and the ash I minimal - requiring a monthly check and dump. I know that this is far removed from the RMH design, but even with electronic controls and monitoring (and fan assisted air feeds and flue control) we still get ash and a warm chimney. If the problem is balancing steady-state energy extraction from the flue versus initial start-up effectiveness, then yes - I agree that some initial burn design feature should be used, just as the Okofen boiler has controls in the initial ignition cycle.
... gosh, that was a long post! I hope it provides some food for thought!