Mark,

I've often thought it would be a good idea to have a window to see what is going on - if only to see if the flame is the right colour. I have tried this with a borax glass window replacing the usual pilot spyhole - I must say it very soon got sooted up and also soon cracked. Do let us know how you get on. I look forward to seeing your video results on this site then maybe we can finally understand the howl.

I've found a .395 jet size works fine with 120psi. As SSSteamer says just a little howl when it's cold and no obvious smoke when it is on full fire.

Mike

Hi All,

The mystery of burner howl is an interesting subject with many answers put forward over the years, there seems to be no rhyme non reason to what will prompt the noise, so many things seem to influence generation of the howl.

The Brooks requires jet sizes of at least .046 with a running fuel pressure of 100psi. I have tried jetting to .050 – 100psi, but this would cause smoke if you are too aggressive with the firing valve when the burner has cooled after a coffee stop or when firing from cold. (Burner plate is a drilled .050 – 20” diameter)

What I find is the real mystery. With a fixed jet size of say .046 – 100psi, the burner will have a deep moan when first fired up lasting a short while, it then goes quiet for several minuets when a much higher pitched howl will start, this will sometime jump one maybe two octaves and as things get hot it will fade away, and you may run the rest of the day, and only hear an occasional cheep on burner re-light. The following time the car is taken out the burner will exhibit the same noises during fire-up but the final howl pitch will not die away, and will stay with you for the rest of the day, reappearing every time the burner re-lights, after a time the noise becomes very annoying. The day after that you fire again, and noise will be the same as the first day, but when out on the road the burner will decide to make a rather nice low volume worbling sound.

Changing jet size, fuel type (28 sec or Paraffin) or mixture (Diesel/Petrol) may influence noise or not, the amount of moisture in the air seems to play a roll.

The 30hp Stanley which as you know I have not had for very long, arrived fitted with .046 – 160psi with signs of poor combustion, but kept good steam, I have reduced the jet size to .038 – 160psi and am getting slightly better steaming, and very much cleaner, the fuel has been varied using the above types, (Burner plate is a drilled .050 – 24” diameter) the amazing thing is no noise, the only howl is heard during fire up, and you might hear a song on burner re-light on the road. The question is, why on this car do all these changes not promote howling?

All the best, Jeff.

Hello Mike and everyone,

This is a good topic for stirring up dialogue! I hope you’re all sitting comfortably, as this is a bit of a long one…

Initially, I will play around with jet size as it’s relatively easy to do, however if I can find a suitable material for a small window I’ll give it a go. Tucked away in the back of my workshop I think I have a small mica window from an oil spray burner, which I may be able to use. This’ll probably take me a while to ‘fit in’ with everything else!

In a ‘previous life’ I spent literally thousands of hours developing a miniature gas turbine engine for model aircraft with great results, we put it in production in 1997. This tiny engine also burned kerosene and developing the combustion chamber was by far the most difficult part. I was always battling with keeping gas flow speeds below the flame front velocity to stabilise combustion – all in a combustor 90mm in diameter and 100mm long – as a consequence, I have an unhealthy relationship with flame velocity where kerosene is concerned! Seriously though, it’s worth noting that the flame velocity in kerosene is surprisingly slow and heavily affected by mixture. For me as an amateur, actual flame velocities were notoriously difficult to measure, because the rapid expansion after ignition changes everything! However I did find that a mixture with maximum richness to support combustion has a flame velocity around 30% the velocity found at optimum fuel air ratio (this is when operating near atmospheric pressure and sensible temperatures). A lesser known effect is that the flame velocity drops sharply again as the mixture is further leaned; the highest velocities observed at optimum fuel air ratio. Flame velocity is incredibly variable as it is affected by mixture, fuel grade, temperature, pressure, atomisation vs vaporisation, swirl, etc..

So – here’s my hypothesis:
If our Stanley vaporising burner is running a little on the rich side (which is no bad thing for flame stability in one sense) the flame velocity may be quite slow; slow enough to be outrun by the mixture coming through the holes in the burner plate. The flame lifts away creating a finite volume of rich mixture just above the burner plate which, having slowed down, then combusts in a little explosion. One of the effects of this little explosion is to briefly push the fresh incoming mixture in the burner plate holes the wrong way, which means that a finite amount of time passes before normal flow resumes, the flame returns to the burner plate and the cycle is repeated. The howl that we hear may be all these little explosions strung together.

So, by bringing the mixture nearer correct stoichiometry, we can increase the flame velocity, hopefully so it now exceeds the flow speeds in the burner plate holes and the flame remains neatly pinned to the burner plate. If this theory is correct, then it means that an over-lean burner will howl also but probably at a different pitch…

It’s worth noting that even at the correct mixture, the gas speeds in the burner plate holes may still exceed the flame velocity, in which case another remedy will be required. I suspect that altering the hole sizes in the burner plate could lead to disaster, as this would change the pressure differential across the plate, change the back pressure below the plate and result in an unwanted fuel/air ratio change (but for different reasons!).

Of course, I could be completely wrong about all of this!

I think that’ll do for now. Looking forward to the weekend for a little tinkering…

Mark Drake

Mark that makes sense. I've seen the model jet engines in action - amazing. You obviously have the right background to solve the howling puzzle - I await further instalments.

I have mentioned before that I have a gas fired workshop heater which has a venturi fed gas burner above which are vertical heat exchanger tubes just like the Stanley boiler. This also howls loudly when first lit but quietens off when warmed up. The howl is much reduced if I open the casing below the burner which presumably increases the airflow, introducing extra air to the outside of the flame and possibly also a better flow to the venturi intake.

I must say I can't see any flickering but the frequency is no doubt too high. your camera would work very well on this heater as access is so easy.

Mike

Does one burner howl?
If the frequency of the oscillations is ultrasonic one burner would not be heard, however two burners would produce an audible beat frequency, the difference between two ultrasonics. As I understand it high pressure jets do produce high notes, or is my memory as muddled as my physics.
If one burner does howl please disregard the above ramblings

Regards Tim 16/3/07

Mike,

Most interesting about your workshop heater – it seems that maybe the warming up process is affecting the resonance / acoustic damping behaviour of the burner – both temperature and density are changing, just some of the many variables which are involved! Does your heater stay at the same pitch, but quieten off; or does the pitch change and then disappear as the temperature rises? Is the heater more prone to howl on a really damp day? I’d be most interested to know.

Re. the camera, it belongs to one of the research departments at work, so unfortunately I’m not allowed to take it off site; but they’re ok about me doing the observations on my burner on site (still in the car of course!).

Sadly, rain, or rather snow, stopped play on Sunday which is when I had planned to try out the new fuel jets; so that bit of work will have to wait. I’m taking the 735 to the upholsterers in two weeks time for a full upholstery refit so I hope that I can at least have a brief run before then, as it’ll be there a while. I do however, have a 1922 model 740 Roadster in which I am installing a new Bourdon 30 hp boiler, so I think I may install a little window in that burner pan, chances are it’ll be a quiet burner!

Tim – now that’s an intriguing idea – it’s perfectly plausible that this could happen. With an oscilloscope and suitable microphone it might just be possible to pick out any higher frequencies, if they’re sufficiently large enough to be identified over the noise that’s bound to be present.

I’ve noticed with the 735 burner that the pitch is always exactly the same (to my ear anyway), if I increase the burn rate enough it jumps an octave suggesting that the audible tone is probably the result of system resonance. If it was the product of beat between two close frequencies, would it be more likely that the audible frequency would wander with changes rather than jump?

Back to my ‘mixture driven’ theory – am I right in thinking that folks with stack blowers find that a slight application of blower stops the howling? I don’t have a stack blower so I can’t try this, I thought I may make a simple compressed air stack blower to fit in the smokehood lid hole just to try it… Maybe a little stack draught just leans the mixture enough to increase flame velocity to keep it close and stable to the burner plate?

Hope you’re all enjoying the banter!

Mark

Hi Mark,

The stack blower does not make much difference to the howl in my experience, but I have found that during firing of the burner from cold, and you feel the time has come when it is safe to close the smoke box flap, if the howl is present at that time, but weak, it will disappear as you close the flap, I suppose changing the resonant frequency of the box makes this happen.



all the best, Jeff.

Mark,

My workshop heater pretty much keeps to the same pitch, the volume of howl starts loud and just declines as it gets warmed up. It only howls for a minute or so and I have not been organised enough to work out whether the noise comes from the venturi, the burner or the heat exchanger - I'll try to find out. Hopefully not much damp air gets into my workshop so I can't comment on the effect of humidity!

On my Model H I find the stack blower can always kill the howl. I suspect the overall airflow resistance of the total gas system from venturi to exhaust flue has an effect on the howl. I think the outlet from the smokebox to the stack downpipe on the H is fairly restricting which is probably why the stack blower is so effective. Like Jeff I find closing the smokebox flap reduces the howl when firing up although I can't think why it should - if increasing the gas flow with the stack blower stops the howl then closing the flap should make it howl more not less. Perhaps like Jeff said this is just messing with the accoustic chamber.

I'm not too sure about Tim's suggestion of the blending of ultrasound to cause audible howls. At one time I had the pleasure of flying in Russian twin engined turboprop aircraft which had a pronounced vibration due to a lack of synchronisation between the propellors. The vibration (known technically as heterodyne) seemed to move from one side of the plane to the other. I don't think a combination effect as Tim suggest would be consistent enough to make the rather pure tone that we hear.

There was a weblink on the SACA website a couple of months ago which described oscillations occuring when gas was forced though a chamber which at the time led me to think that this was the explanation. www.raczynski.com/pn/fluids

Your flame front theory has now confused me again.

MIke

Hi Jeff, Mike,

Thanks for your input, I’ve got some interesting stuff to report, another long one…

I tend to agree with the you, Mike, that the ultrasonic / beat idea is unlikely; the tone just doesn’t have that kind of quality and also tends to be fixed in it’s fundamental and harmonic frequencies. The beat effect in prop aircraft is quite well known I think and can be extremely damaging.

Both your comments about smokebox flap and blower are most interesting – I too have had similar experiences with the smokebox flap on first start up.

However, down to the good stuff-

I took yesterday off to finish a plumbing job in the house. The job went so well that I spent the afternoon mucking about with fuel jets, here’s what I found (all tests initially with fuel at 140psi and ambient temp 5°C).

Jet Size 0.032” – Struggled to get the burner to light-off on hexane and flat refused to light on kerosene at all. The mixture was so lean that I could see that the pilot flame wouldn’t propagate across the burner – if I partly covered one of the venturis, I could get the burner to light off rather badly. Not recommended! But at least it showed that the mixture was too lean to burn.

Jet size 0.035” – Burner would just run on hexane, but the observed flame was a very pale blue in colour, unstable and not covering the burner plate completely. The burner would not run on kerosene, but would briefly light-off when fuel was first admitted.

Jet size 0.037” – Runs reasonably on hexane, but clearly unhappy. When switching to kerosene the burner would light-off, the flame was a thin blue in colour with lilac near the burner plate but I could see that flame did not entirely cover the burner plate. There was also a significant amount of unburned fuel vapour issuing from the smokebox flap.

Jet size 0.0385” – Goodness, what a difference! The burner lit-off sharply on both hexane and kerosene. No noticeable unburned vapour coming out of the flap / stack. The flame had a bright blue colour and sat fairly neatly on the burner plate. On first lighting, the burner howled a little but this disappeared within a few seconds. Once everything was thoroughly warmed through and I had about 300 psi of steam, I found that I could run the burner at full fire with no howling at all – something I just couldn’t do with the 0.049” jets. The burner just produced the usual strong hiss, with a quiet rumble.

I took the car for a run locally to see what the steaming was like, and this is where things get really weird. As soon as I moved off, the burner began to howl, literally within a ˝ turn of the wheels! Stop and it disappeared again. On the move, however, the burner still howls, but much less intensely, with a warbling quality, like Jeff describes on his Brooks. It also comes and goes quite readily, suggesting that I’m right on the margin between howling and not. It was a very cold day, so I’m looking forward to trying it when it’s warmer. Also this jet size gave much sharper burner response on the road, particularly when lighting-off when opening the throttle at the bottom of a hill, for example.

In reflection, I would say that fine tuning the mixture certainly has merit – just 0.0015” made a massive difference. However, I feel that burner howl is the product of multiple effects, a strong contender is the resonant behaviour of the whole system geometry. I think I’ll have a look at geometry and damping next.

Could temperature difference of the superheater cause the burner to only howl on the move and not stationary? I’m a bit stumped by that observation – I’m joining the ranks of the confused too…

Clocks change to summer time next weekend = more steaming time!

Regards all,

Mark

Mark,

Very thorough experimentation! How can you see the flame - have you already installed your window? I can't see my fire at all but would very much like to.

I think the sudden appearance of howl as you start to move must be due to changes in the intake and flue outlet gas pressures due to movement. I once put a small shield about 50mm fowards of the branch forks (front intakes on non condenser car) and this caused major howling even when standing still.

When you said temperature difference of superheater did you mean vaporiser? I think there is a connection between the adequacy of fuel vaporisation and howling as a fuel spray in droplet form probably kicks in less air than a hot gas stream thereby changing the mixture - on top of this, imperfectly vaporised fuel can probably flow through the jets at a faster rate than pure gas. Who know what this does to the mixture but it may explain why a burner will howl when firing up but not when warmed up, and the litte chirp of howl which happens if the car is stopped and then moves off while the burner relights. I doubt there is much variation in vaporiser temperature just from moving off.

If you consider the area of the jets it is not so surprising that a 0.0385 is detectably different from 0.037 - the area is 8% greater.

We've got a massive dose of salt on the roads again so can't play at the moment.

Mike