I was going to post this in the rapidly heading off-topic flywheel thread, but I'll post it as a new topic. Feel free to put your close calls in here. Here is my little story of my most recent close-call. It's a bit lengthy, but it'll give you something to read over Christmas.

Merry Christmas 

---

One of the more important things to have in hydraulic systems is tamper-proof relief valves. In things like forklifts and loaders, you set them to open at a pressure that equates to a little bit above the max SWL and your machine simply will not lift anything that it's not safely capable of. Hence the reason that they have to be tamper-proof. 

Found a loader this week that had a problem with it's engine stalling when in remote control. I'm an auto electrician, so the words "loader" and "remote control" and "problem" means that it's now my problem to sort out and the diesel fitters flee in the other direction as rapidly as possible. In fact , any mention of control problems on this type of machine and they pass the buck to me, as it has a particularly maddening electro-hydraulic pilot control system, which is controlled via a CANbus system, with a little module reading the joystick position, sending a little data stream to another module about 6 inches away, which mixes that with a bunch of conditions (and possibly a bit of black voodoo magic, it's hard to tell from the manual) that all need to be met before it puts out a PWM output for the control coils which operate the pilot valves , which operate the main shuttle valves, which actually move the damn boom and bucket. Wouldn't want to make it too simple or anything. So yeah, the fitters flee when one of these things break down. Bastards. But I get 'em back every time I find a big , oily hydraulic leak, so I suppose it all evens out in the end.

So anyway, I start it up in remote with the little joystick pack, flick it to "operate" and sure enough , after about 5 seconds the idling engine loads up and stalls out. The operator standing next to me tells me he just hits the "full throttle" button on the sticks and carries on regardless, it works fine at full noise. It's apparently been doing that for about 3 weeks, but Operations didn't want to give the machine to Maintenance as they "really needed it". Well, that's what they always say, so there's no real surprise there.

The only thing different between standby and operate in remote is that a little coil on the bucket pump has its current reduced, making the variable displacement pump - about the size of a small car engine - go to full stroke and hence max output. In another particularly fine bit of design work, if you burn that coil out, the pump loads up all the time and the next time you turn the engine off , it can't start up again due to the pump load. Did I mention that the coil runs all the time - even with the ignition off, and it gets hot enough in normal use to blister you? And the fact that they always - always! - fail when the machines parked in a puddle or a pile of mud'n'crap? And that you can't reach the coil without climbing under the machine? Excellent design there guys. Real good.

So, I think "Hmm. Must be something to do with that coil." Clamber under the machine in all the mud and crap, swear at the manufacturer for putting the coil in such a stupid spot and check the coil.... it's ok. Swear a bit more , because now it means it's something else out of the ordinary. Dammit.

I get back out and hop in the cab to check out the machine's CANbus system... everything really looks to be normal on the control/electrical side of things. I've got the covers off in the middle of the machine and when I load it up in remote with the little joystick pack on my lap in the cab, the pump hoses - which are nearly 3 inches in diameter - twist and flex about a good 6 inches back and forth. I can also hear this little funny metallic rattling noise coming from the front of the machine, which I realise after about 30 seconds to be the 330 bar shock relief valve in the implement control block at the front of the machine going off.

"Hmmm," I think. I look up the specs on the pump and double-check what it's meant to be -  280 bar.

So I put a 350 bar gauge on the outlet of the pump, hang the gauge in view of the cab and start it up. It winds off the scale. I turn it off pretty smartly. I go get a bigger gauge, with which I find the pump pressure peaking out at a lovely 375 bar. So I shut it off, clamber under the machine again (dammit!) and find the pressure adjustment on the pump and wind it out about a good 6 turns. This gives me 240 bar, so I wind it in a bit and get the 280 bar I was aiming for.

The pump has been running a hundred bar over the rated maximum for the last three weeks. That's 1500PSI higher than spec, or about 5500 psi all up for you imperial folk. So, I go back and tell my supervisor about it all. It's the end of the shift - a night shift, so it's just on 4:30am. He tells his supervisor, who happened to be on the phone at the time with him doing the end of shift handover. He gets off the phone with my supervisor and rings the maintenance superintendent, who nearly has a heart attack.

I find this out about an hour later, when I find the superintendent in my little hideout, just as I'm about to pack up and go home. After about 10 minutes of discussion, we both go to the machine on the job, throw the operator off it, place a "Do Not Operate" tag on it and take the key with us.

I go home , to sleep. I come back to work the next night to find:

- The machine in the workshop, in a fairly-dismantled state.
- The bucket pump, out on the floor with it's control valve very carefully dismantled on a bench.
- The entire length of line from the bucket pump to the implement valve stretched out along the floor next to the machine, with all it's fittings , etc attached. Basically it's the 280 bar circuit, complete.
- A lot of section of that line spray painted white. A few fittings as well.

Turns out that the bits that are painted white are spots where the hose has started to fail - there's blisters under the outer lining. The fittings that are painted white are cracked. The hose guy has stripped back sections of the outer hose liner and you can see the braid is stretched/broken in spots. He says that 20 to 50 more hours or so on those hoses and they probably would have let go.  The cracked fittings could have let go at any time. The rep from the pump manufacturer has very carefully dismantled and measured all the bits in the pump control valve and says that they're all within spec and that the position of the setscrew where I adjusted it to 280 bar is close to the normal location for that pressure, so it's not like a spring has collapsed or something.

Now if one of those hoses develops a pinhole leak at 375 bar , it's enough to puff you up like a balloon with burning hot oil if you get in it's way. The hoses are about 25 feet long, basically unsleeved/uncovered, chock full of tee pieces, valves, joins and one section passes through the articulation at the middle of the machine, right next to the cab door. Maintenance people test the machine with the engine running, with the top covers off, standing right next to those hoses as they snake their way from one end of the machine to the other.

So at this point, the shit has hit the fan. The mine manager wants to know what's going on, the state government dept that looks after underground mines would like to know as well, thanks. I write out what I found and did on the incident report. The other 4 loaders of the same type are immediately stopped and get their pump pressures tested, they're fine.

The only real conclusion is that "someone" has adjusted the bucket pump to the higher pressure. Someone about three weeks ago. A quick skim through all the work orders for that machine reveals that there were a couple of people from the cross-shift that went to look at that machine for complaints of "not being able to dig/lift." It was my last night shift last night, so they're due back today.

It looks pretty much like someone will be taking some unexpected time off after Christmas.

I've searched all over and while I can find plenty of topics that mention safety in passing , the only real one that even begins to present the risks involved is Guy_Fawke's "How to kill yourself with a listeroid", and that sort of went to pieces pretty quick.

But it's something that needs to be addressed. There are a lot of people - like me - that are new to the game. There are a lot of people who have - after years of working around listers - become a bit jaded about the safety issues. Please, if you think I'm preaching to the choir, don't say anything at all here unless you're here to help. I'm really trying to keep this one focused.

Quite a few posts about the place seem to hint that people have never done anything like this type of risk analysis before. I want you guys to post any questions you've got as we go along.

So, I'll start the ball rolling. This is a thread that will be used to help you identify hazards and courses of action to reduce the likelyhood and consequences of them. I'll work over the common hazards, but by no means is this the definitive list of hazards on a machine. You're machine will vary from mine, and you might have some nast hazards that I don't. So keep an open mind , and an eye out.

I've lifted bits of this verbatim from our training manual here at work, so take that in mind when it talks about "Business interruption"

Ok, what the heck is Risk Management?
It's the process of

1.    Identifying hazards in the work area
2.     Assessing the risk posed by them
3.     Implementing effective controls to reduce the risk posed by them
4.     Reviewing the effectiveness of the controls on a continual basis

Ok, so what the heck is a hazard?

A hazard is something that has the potential to cause injury, damage to equipment, environmental impact or business interruption.

Ooook, so what the heck is a control?

A control is something that reduces the exposure to the hazard. On the one extreme, you can get rid of the hazard completely by some method, that's the most effective control . On the other extreme, you could put up a sign that says "anyone going near those spinning flywheels will be reprimanded". That's a poor control because it depends on the person. And you can never depend on the person.

I'll now post this:


Normal copy here: http://listerengine.com/coppermine/albums/userpics/10075/risk%20card.jpg
Yes, I lifted it from work and yes, I have permission from our safety manager there to use this image.

This card can be used not only for risk of injury, but machinery damage and environmental risks as well.
Now, how to control all these hazards that we've identified as big risks?
Use this :

Normal copy here: http://listerengine.com/coppermine/albums/userpics/10075/Controls.jpg

This lists the controls from most effective to least effective. I'll give examples:
Eliminate: The best choice, obviously. Remove the hazard... no hazard.
Substitution: Can we do something some other way, or can we get a part that doesn't have that hazard with it?
Engineering/Isolation: Engineering - Making something strong enough that it doesn't fly off and come to you. Isolation- Keep you away from the moving bits by guards.
Administration/Training: Procedures to do something, training on how to do it.
Personal Protective Equipment: Safety hats, gloves, glasses , aprons, earmuffs.
Behaviour Management: "Keep away , or you'll be reprimanded"

Now, posting the list of hazards from Guy_Fawke's thread, and a few more:

1) Large stored energy in rapidly moving open flywheels.
2) Massive and top heavy. Obvious handling issues, in whole, or in parts.
3) Exposed Belts.
4) AC power.
5) Hot bits. Coolant, Oil, Exhaust
6) Bits that pinch , as opposed to rip off. Exposed valve train, internal gears and parts.
7) Stored fuel in the area, fuel near hot surfaces. Inevitable leaks. Risk of fire.
8 ) Exhaust issues. Noisy and dangerous gasses.
9) High pressure from injection pump.

The way we do it at work is two-fold :
 - an assessment method where we just look at the whole machine and see what the dangers are. I'd use that list above as a start.
 - an assessment method that's task-based.

I'll use the task-base method as an example.

So today's task to work on reducing the hazards is starting the machine.

What can go wrong? We'll go through the steps in the task.
Reduce each step down to a single action or actions in a similar area. Don't combine actions (eg. setting the fuel rack/decompression and winding it over). Setting the fuel rack/decompression can be considered as a "Setting up for cranking" action as they're in a similar place on the machine.

Ok, Step 1 in starting the engine:

1. I grab the crank-handle and walk to the machine.

STOP. Have a look at your surroundings? Is it dark? Wet? Full of junk? My areas well lit and dry, but it's full of junk.

Could I trip over some piece of junk in my cluttered workshop? Yes. That's a hazard then.
So I get out my little chart and work out a risk score.

Likelyhood? For my place, about once a month. There's crap everywhere. I look up the chart, and thats a "B"

Consequence? I go arse-up and knock my head. That could be a lost time injury - a "3" (ie. Hospital and a day off work)

Risk score? Where B and 3 intersect - 17.

That's a high risk. What to do about it?  Looking at the control pyramid, the top one is to eliminate the hazard ie. tidy up the damn shop. I could use some of the other methods from the pyramid, but eliminating is the best one as there's no hazard after that. I could wear a crash helemet (personal protective equipment) , but that doesn't reduce the likelyhood of it happening, only the consequence that I'll get hurt.

So anyway, I tidy the shop. An hour later and the place is a lot better. Now to double check.

Likelyhood? For my tidied-up place, it's unlikely. So I'll make it "D".

Consequence? I could still hit my head. If I lined the place with pillows or wore a crash helmet, the consqeuence would be reduced. Don't get fooled on this bit. I've cleaned the workshop and reduced the likelyhood of going arse up, but the consequence of me going arse up is still there - a sore head. Still a "3" then.

Now people have differing ideas about likelyhood and consquence, but it's really up to you. You should really do this with a couple of other people to get a consensus, and to stop you becoming blinded to hazards.


Looking that up again gives me a risk score of 9 - it's been reduced to a moderate risk.
So I've reduced the risk of walking over to the machine. Good. That's a start.

I'm going to pause here, because I've got things to do elsewhere, but I'll leave the next step in the task out in the open, if anyone wants to give this a go :

Step 2. Checking fuel/oil/water and engine settings (eg fuel and decompression).
What are the hazards associated with those tasks, what is each hazard's risk score and what can you do to reduce that risk score?

Alright, I've started this topic with the goal of designing a steel frame, with isolators of some sort to prevent vibration from being sent from the engine to the Rest Of The World.

First up :
Please leave all the arguments about why in the other threads about the place. By all means, if you've got something that to say about how to properly isolate the engine with resilient mounts, speak up. If you wish to discuss alternatives to the steel frame/resilient mount, or wish to debate general engine life running on mounts, take it elsewhere.
Specific cases of engine longevity issues (eg. damage due to undampened resonance at particular frequencies) are welcomed, as long as you are willing to help the attempt to come up with a solution. A solution that doesn't involve concrete. 


Might be handy to skim through a vibration isolation primer to get a quick idea of the concepts. It's at :
http://www.wtc.net.cn/Primer_Vibr_Isol.pdf

Ok, to start off:

You've got an engine that's been relatively well balanced. It doesn't chase you around the shed when its bolted to a couple of wooden railway sleepers. Much.

Pretend that you live in a treehouse, or a small boat. You need the engine, there's no alternative to a 6/1 and an extra 2 tons of concrete is not physically possible.

Also pretend that your engine is mounted next to and powers an extremely sensitive set of scales that you need to be as wobble free as possible.

Don't poo-pooh these two requirements, I've been in a situation very close to it 

Let's discuss the general specs and requirements:

- A listeroid 6/1 running at 600-650RPM with two main forcing frequencies, a 5Hz power pulse from firing once every two revolutions and a 10Hz pulse from flywheel imbalance/reciprocating forces.

- It weighs approximately 500kg, once you take into account a decent frame and possibly a generator as well.

- You want the maximum amount of vibration isolation.

- You'd prefer if it didn't oscillate a great deal while sitting on those mounts. I'll arbitrarily set a limit of 10mm worst-case. More is briefly allowable on spin-up/down.

From all this, we should be able to design a 'standard' resilient mount that reduces vibration well and can be made relatively easily with reasonably available bits. It should have enough info about it that you can make an informed choice as to substitutes or alterations.

So, what would you use as the resilient mount? Frame design? I've got a pretty good general idea - see my "6/1 and generator setup" thread, amongst all the arguments. I'm interested to hear what other people would use. Sucesses, failures, etc.
Outline your resilient mount setup - what it's made of, how well it works. If you have an idea of why it worked or not, let us know and we'll attempt to back it up with the proper calcs as to why it was good or bad.

I'm slowly working my way towards getting a jkson 6/1 and a 5kW ST head for it. It'll be used in an off-grid setup to top up batteries once a week or so, with solar making up the most of the charging capacity. Charge time from flat of the battery storage should be about 8 to 10 hours of full load (3kW) on the 6/1, but I don't expect it to be completely flat all the time... just half the year in monsoon season  . I'm using new nickel-iron cells as - while inefficient - they're immune to discharge damage. I expect the system will be used and abused lots.

Anyway, going through the math involved it seems I need a 250mm diameter pulley on the generator to get 1500RPM and 50Hz at aroundabout 620-650rpm. I'm not completely sure of the flywheel size of the jkson engines - 23.5" ?

So, rough gameplan is :

- Source a 250mm, 8 groove poly-v pulley with a bush to suit the 38mm shaft

- Find a suitably-sized belt and bung the 6/1 and genny together on a solid frame

- Possibly get a small tensioner on the frame as opposed to stuffing around with a slide of some sort for the genny, as I'm sure to screw that up. - the pulley mob also stock some nice stepped tensioners that also serve as belt trackers.

- Mount frame with fairly solid rubber mounts bolted to a concrete pad. (Fairly solid as in rated at 500kg each rubber mount. It won't be all that flexy. Hopefully)

- Run cooling pipes to a fairly thin tall rectangular cooling tank and let it thermosyphon away. Dimensions guessed at 100cm long by 150cm high x 10cm wide or so - just want plenty of surface area for convective cooling, right? Anyone good with the calcs for this? Ambient is about 20 - 30 degrees C year round. There'll be a heat exchanger in there somewhere to boost my solar hot water system, as when it's cloudy I'll be needing to run the genny to charge the system up anyway.

Anyone see any showstoppers in this?