the use of #3 by anyone less than a degree'd engineer is probably being irresponcible, a path the average DIY'er should not even entertain. The shear numbers of variables envolved coupled with the variable nature of the engine metallurgy, machining, assembly etc. make for such a complex problem that there is likely no way that the average DIY'er will be successful unless he is very lucky.

In this instance I would definitely recommend the assistance of people who sell the suspension components. As I've mentioned before, while the listeroid world is firmly stuck in the non-resilient camp, the maths and engineering for resilient mounting has already been done for many , many similarly large pieces of reciprocating and rotating equipment. It's not a trivial exercise for the person who has never done it before, but for the supplier of such items , well, they do this every day. And they do it for equipment much more expensive than your $1000 listeroid, where the stakes are much, much higher (downtime or equipment damage can run into millions of dollars) . No engineer worth his salt these days mounts a compressor, motor, centrifuge, or pump directly on concrete.

I've done the math roughly. It was basically to check what the suppliers of airmount systems told me, because I like to suss things out for myself. In the 5 minutes on the phone with an airmount supplier, he queried the mass of the system, it's RPM, consulted a few charts and gave me the answer that took me about 2 weeks of research to agree with. In passing, he mentioned all the things we've gone over here - constraining the movement to one plane as much as possible, good balance, checking for rocking action, snubbers just outside the normal range of motion, etc.

And one more thing - the math suggests a very poor amount of isolation with mostly rigid rubber mounts. It also hints of the chance that you'll be unlucky enough to find your mount resonates at some RPM you don't want (eg. 650RPM, or in an overrev situation). All suppliers of engine mounts should be able to tell you what that RPM will be, or provide you with enough information that you can work this resonant RPM out yourself in 5 minutes with a pocket calculator.

Make sure you do it, for your own safety mainly, but also so we don't get guy_f back in here crowing about how he told us it would all go horribly wrong. I don't think I could handle that.

CO2 fire extinguisher plumbed into the intake manifold.

As long as the thing spins down in time before the CO2 runs out......

Tried murphy switchgauges bob? Added advantage is that trip points can be set.

See : http://www.fwmurphy.co.uk/products/electric_gauge/electric_eg.htm

You can also get pure mechanical ones as well which could be useful.

http://www.fwmurphy.co.uk/products/pressure_vacuum/pres_20p_25p.htm

A little pricey, but very, very reliable.

(Hey andy - that picture link goes nowhere for me - is it something similar to this?)

I was thinking with my resiliently mounted engine, that a simple air shutoff could be built as follows :

- Place a relatively soft rubber ball inside the air cleaner, above the intake, on something similar to a golf ball tee, or a teaspoon.
- When engine gets out of control and starts shaking on it's resilient mount too much, the ball falls off the tee and drops down to block the intake.

The size of the tee determines how much shaking does it - obviously you'd want it to be somewhere a little bit than your average startup/shutdown jiggles to avoid nuisance tripping. For a manual shutoff, you could weld the tee/teaspoon to a rod that goes outside your air cleaner and all you have to do is rotate it 90 degrees for the ball to fall off and shut the engine off.

Resetting it requires you open the air cleaner housing and fish the ball out, but that's nothing too serious. If need be, you could attach a bit of string to the ball and run it out of the housing directly above the tee, so that pulling on it would lift the ball back out above the tee and then lowering it would sit it back on.

Sorry, saw an opportunity before and couldn't resist.

Anyway.

A chain drive with a spring tensioner idler to keep the slap down a bit is do-able. I don't know how it would go with the listeroid's torque/speed variations.  At light loads, there might be occasions (eg, coming up on compression) where the generator's inertia makes it want to drive the engine - you'd get a bit of chain slap there. Fully loaded there should be enough generator drag to ensure a pretty decent tension on the drive side of the chain regardless of variations in engine speed. There's a tradeoff between sprocket size and chain tension there as well. You want good tension on the chain to help prevent slap, but too small a sprocket puts a big load on the chain leading to shorter life.

It's really one of those 'try it and see' things - there's no spectacular hurdles to overcome really, so it'd be interesting to see how things go for you.

Cogged drive belts are pretty much the pinnacle of efficiency , but saying that, they're only a few percent above chain drives and for the outlay it might not be worth it. But they are significantly quieter than chain drives if that becomes an issue.

There shouldn't be any twisting forces escaping from the engine-generator frame - all the torque/countertorque forces from generator load are contained in it. So as long as that frame is suitably rigid, there should be no issue with gyro forces from the generator load. Gyro forces from rocking introduced by unequal flywheel masses is still there. Which was why I was considering a mount with a (rather strong) hinge at one end, basically restraining the resilient mount to the vertical axis.

With the balance rings - I've a digital video camera with a high speed shutter. I plan to put a loop of clear hose with a bubble of coloured water/honey around the flywheel and film it. If you play it back frame by frame, you should be able to see where the bubble sits on the flywheel - that's where you need to add mass.

The water and plastic hose is light enough for me to be reasonably unconcerned about it flying off should that occur, as opposed to magnets/lead shot/etc.

Ok, so apart from Guy_F's tangential rant - where he makes the usual load of sweeping generalisations and gives very little hard evidence to the topic at hand - I've seen a couple of good responses.  I've poked through the board, and whilst I've seen many,many references to what-may-hap? , I see very few references to actual injury or death directly attributable to a lister/listeroid engine failure of any kind. I say again, mainly for Guy_F's benefit -

- There are no doubt quite a few people who own engines here.
- They are all of extremely variable quality. (The engines, that is!)
- Not knowing precisely how many people run their machines 24/7 , one can only guess at the total hours run by all of them. 200K?
- I have yet to read a post where someone has been seriously injured (that is, hospitalised) by their engine. I've seen a lot of 'could haves', but nothing yet that's gone all the way to the final result. Apart from dumb luck (eg. it exploded but there was no-one around) - all the other things that stopped it from reaching the point of injury are legitamate controls, such as monitoring machine noises, good shutdown controls, guarding/retaining parts, regular inspections, co-incidental inspections (saw something while glancing over the machine) etc.

So, are we blowing the probability of it happening all out of proportion? There's no doubt that there is some finite level of risk involved. Whether it's the bogeyman that some paint it to be is debatable.

Don't give me references to propane tanks, shuttles and a million other things that have bitten people on the ass that are 'sort of the same' as running a listeroid. There are enough people on this board that we can get a good idea of the 'average' listeroid's safety in 'average' conditions. For a specific engine - we'll never know. Your engine might spit its flywheel out tomorrow and there's not a damn thing anyone here can do about that, that part is up to you to manage as best you can. To really manage that, you need to know the probability and the consequence. We know the consequence, we don't know the probability. Without the probability you can't tell if it's a big risk or if its something that you don't lose sleep over.

So post your engine failures. In addition to that, if you've done some decent time on an engine and everything's been fine, post that too.