Bristol Mercury revisited

[Admiral_dk]

No problem Mike.

I have been working at work for the duration, with a total 14 days of holiday - so my reality has been completely different from a good number of other members here.
But I do have not problem with the need for a change of direction (though the options are somewhat limited in these times) ....

What ever you decide to spend your time on Mike - I hope you enjoy it 

[AVTUR]

Mike

Good luck with your future ventures and I hope you continue with the Seal.

Maybe we will meet at RRHT.

AVTUR

[Vixen]

Hello AVTUR

I believe Per may have got hold of the wrong end of the stick. I am not about to say goodbye, I am not about to disappear. My intention was only to cut back on the time spent in the workshop each day. I plan to continue to contribute to this excellent forum from time to time. And when conditions permit to return to the model exhibition scene with renewed energy.

Do you remember that caption on the television, when we were youngsters,     

"Normal service will be resumed as soon as possible"

In the meantime, I still have lots of modeling work ahead of me. The update program on the two Bristol Mercury engines has be completed

That still leaves the Bristol Jupiter     http://www.modelenginemaker.com/index.php/topic,7559.0.html

The Mercedes Benz W165      http://www.modelenginemaker.com/index.php/topic,5142.0.html

And the Westbury Seal     http://www.modelenginemaker.com/index.php/topic,9709.0.html

All are waiting to be progressed. The Westbury Seal is very near to completion.

Stay in and stay safe

Mike    strictly AVGAS

[Admiral_dk]

To be honest Mike - the way I read your original post, I thought that it could be understood in more than one way and I simply wanted to be sure that if the meaning was about leaving - I would make sure I expressed my gratitude ..... and if not - I felt that you would point out my error 

Per

[petertha]

Work on the combined pressure pump and scavenge pumps has continued. The housings have been cleaned up and all the inlet, outlet fittings, blanking caps and pressure relief valves have been completed and assembled. Mike

Mike, I know this is an older post & you have lots of other interesting projects on the go. But did you ever bench run your oil pump system? If not, no problem. I'm curious if the key dimensions were more influenced by scaling the actual engine pumps or if you also tweaked with model specific design parameters? Any rough idea of delivery pressure, flow rate, gear dimensions, RPM etc.

[Vixen]

Hello petertha,

The oil pressure and scavenge pumps for the Bristol Mercury engine are (nearly) exact copies of the full size oil pumps, hence appear to be more complex than necessary. I'm a slave to scale.

A scavenge pump should have 1.5 to two times the capacity of the pressure pump that supplies the oil to the engine.

Pinion pumps are constant displacement pumps. The amount of oil pumped is therefore dependant on the size of the gear teeth, the number of teeth in the two gears and the rotational speed. True, you will get some leakage over the ends of the teeth so keep the clearances as small as possible. The pinions in the Bristol Mercury pumps are made from 0.8 MOD gears 0.375" (9.5mm) diameter. They rotate a about 7/8 crankshaft speed. it's always good to keep the pump speed as low as possible.

There seems little risk of cavitation at model engine sizes as the pumps are small with comparatively low flow rates.

Output pressure depends on the load. An open ended pipe, feeding ball races, will see very little output pressure, whereas a plain bearing, which restricts oil flow, will increase the pump pressure. The pump pressure is therefore dependant on the restriction placed on it's output. You will be surprised how little oil is require to lube the bearings, remember most two stoke engines are lubricated only by a tiny quantity of oil mixed in with the fuel.

Most full size aero engines and racing engines rely on excess lube oil flow to help cool the internals of the engine, thereby reducing the size and drag of the coolant radiators. Not normally an issue with model engines.

Cheers

Mike

[petertha]

Thanks for the info Mike, very helpful. I was wondering if you had bench run the pump or had a rough idea of numbers by your own design because behind the scenes I was busy plagiarizing & bastardizing & metrifying David's pump spreadsheet for my own selfish purposes. I have been thinking about purchasing metric gears in the similar sub-1.0 module range to build an evaluation pump. Even with small gear diameter & reduced module they appear to kick out a lot of fluid. So much so that I was suspect I messed up the math or something else going on. Your gear width is longer yet (you mentioned scale) so I thought that would be particularly useful info if you filled a X liter bucket in Y minutes. Probably this 'pump design' discussion best belongs in the other more general gear pump post. I didn't want to clog up your build post, so I'll leave that to you. But for interest sake, just eyeballing the gear width & guessing at some other parameters, I get numbers like this. Does it seem sane to you? Happy to share the spreadsheet, particularly if you see a bust.

[steamer]

Peter,

Just a note.    The pressure in my spreadsheet was an INPUT.    I wanted 10 psi available in my pump, and I wanted to know how much power that would consume at that pressure.   The  actual pressure in YOUR system will depend on the down stream plumbing/orifices/restrictions/viscosity/

I am going to balance my fixed flow nozzles to an appropriate flow rate at probably 10 psi....plus or minus a bunch.   Again, what it takes to make it work will not be very much IMHO, so it's a balancing act that I'll need to sort out on the bench, and not a closed form solution.....but the spreadsheet gives you some idea of where your at as a starting point.....

Now what is an appropriate flow rate for a bearing!?   That's a great question!   How much will my system need?    Also a great question!  and one I've been asking myself for some time!


To answer this, I sat an thought about this for sometime?  and I finally decided to perform this thought experiment.

I asked myself, if this bearing was open, and running on the bench, as I stood over it with an oil can, how much oil would I give it as it was running at 8000 rpm?  

Remember now, this is a bronze bearing on a hardened A2 tool steel shaft, quite polished and good fitting.   Method of application is basically splash lubrication with a boundry layer lubrication. meaning it's hydrodynamic.   
The purpose of lubrication is to reduce friction and wear and carry heat away from the bearing.    OK.   

Wow...not very scientific I know.    Guilty, but splash lubrication is by definition not very scientific.....

Furthermore....remember that the 4 stroke model airplane engines on the market lubricate with nothing but adding oil to the fuel and count on some blow by to lubricate the VERY simple bearings, ....so how much oil is transferred to the bottom end of one of these engines?    NOT A LOT!

So rational me decided I would lubricate this bearing on the bench at 1 drop every 2-3 seconds.   I call this the rational man approach.....  Not scientific, but....

How much oil is in a drop of oil?    Google tells me 0.064 milliliters.
How many bearings am I feeding in this engine?     6 double throws and a gear train ( I am assuming the same flow to the gear train)

1 drop/3 seconds x 7 is works out to 140 drops per minute... 

140 x 0.064 milliliters is 9 milliliters/minute    That is going to be a very small pump!!!!!

Too small!!     These pumps are pretty fussy to make and function well, lots of concentric features that need to line up ect, and small gears are pretty fragile.

My pump, which is about as small a pump as I think I can make reliably.    It is capable,   theoretically to 560 milliliters a minute!!!   Or about 62 times the minimum amount of oil I think I need.... an that is the rub.     I think the pumps we can reliably make are going to be big compared to what we need!

My pumps are 48 DP which is equivalent to 0.529 MOD.    Small!

I also looked at Ron Colonna's Offy plans which if IIRC, is also a 48 DP pump and runs at 1/3 engine speed  ( I'm running faster at 1/2 engine speed)  and got a similar size oil and scavenge pump, and he has silver bearing mains on 4150 steel crank, with drilled crank oil passages.   I have ball bearing mains and bronze bearings running on hardened steel crank pins....which is a great bearing pair.    I'll dig up the build book and quote chapter and verse, but it's a similar size to mine, and it's slightly larger engine displacement IIRC.

so now for me,  the next step is to test with my preferred oil spray bar which will have 7 nozzles aimed at the components i've mentioned.    I'm going to start with a 0.005" hole as a nozzle and 10 psi....ish    but I think I have plenty of leeway here.....I don't think I need very much and I think most of the oil will be cycling back to the oil tank.

Dave

[Vixen]

Hello petertha.

Thank you for your interest in the Bristol Mercury oil pumps and for this lively discussion. I will create a new topic on oil pump design and move this discussion there, so it will be easier for other to reference it later. Leave that to me to do, in the mean time keep posting here.

No, I have not built a flow test rig to measure the flow rate of the pressure and scavenge pumps for the Mercury engine. I am sure the flow rate is amenable to calculation. I have looked at the gear displacement calculation in your spreadsheet and cannot come to terms with it. Your calculated pump flow rate of 2,099 ml/min ( 2 litres/min) at 2000RPM looks to me to be rather optimistic

The oil being pumped is entrained in the gaps between the teeth of the two gears. I would therefore have expected your displacement calculation to have considered both the outside diameter and the depth of the gear teeth, multiplied by an estimate the the tooth to gap fill ratio and the number of active gaps surrounding the two gears.

The pressure pump gears for the 1/4 scale Mercury engine are 9.6mm OD and 9.6mm wide, 0.8 MOD.

As for the required flow rate, most agree that only a small proportion of the flow is used to lubricate the bearings, the majority of the flow is used to cool the inside of the engine. Internal engine cooling is particularly important with air cooled engines, like the Porsche 917 and most radial aircraft engines. The oil flow rate of a full size oil pump will therefore be determined by the engine's worst case operating conditions i.e. take off (or racing) in hot, high altitude, desert conditions. Fortunately, we only need to consider running our model engine in our back yard conditions.

Mike

[steamer]

Peter's reference was from my spreadsheet Mike, as is the volume approximation formula.    Spreadsheet errors aside, which I have not looked at Peter's yet.

As far as cooling, absolutely, lubrication is a lot of cooling for sure.    That said, most model airplane 4 strokes are air cooled and function well with nothing but oil in the fuel/blowby.   These models don't put out the amount of heat that the full size does  ( cube root of size and all)

Dave

[John Roberts]

Mike, I have only been following the forum builds since I started making the Triumph t100 engine. I am totally blown away at the detail that you have put into this engine. I thought I had patience but for you to make two engines, unbelievable. Drawing an engine in 2d can be difficult to spot errors as you know. Your ability to make do with what you have got is commendable.
 It must be difficult to have milled the larger parts with such a small  X , Y axis envelope on your mill.
 I do have 3d modelling skills and have 3d cam. I know little about g code and would not know where to start adapting code to cut the complex parts you make. Thankfully I don't have to.
I have not decided what my next project will be, but I am going to have to up my game to come close to what you produce!
Like you and others, I have em-braised CNC machining and the challenges it brings.

John.

[Vixen]

Thank you John,

Like Cortés, when he discovered the new world; I burned my boats as well; so CNC was the only way forward. The only handwheels I have in my shop are the tailstock handwheels on the two lathes. Everything else is done by the numbers(CNC). I have been using the same CAD package and much the same CAM software for more than 20 years and feel I am at last getting the most from it.

That's what they call experience. And as we all know, experience is that which allows us to make the same silly mistakes every time .... but with total confidence. 

Mike

[Admiral_dk]

That's what they call experience. And as we all know, experience is that which allows us to make the same silly mistakes every time .... but with total confidence. 

I know that one FAR to Well    + it doesn't only apply to CNC 

Per