To swirl or not to swirl

[britishrocket]

Greetings all,

I've not posted on here in over a year now, but I have a question which I think could be answered by the assembled luminaries.

I've been reading a paper entitled "The development of an expendable gas turbine engine" by Dr. D W Artt and Dr. N F Adams, of Queens University, Belfast, United Kingdom. Here it is:-

www.gasturbineworld.co.uk/The%20development%20of%20an%20expendable%20gas%20turbine%20engine.pdf

By "expendable" they meant simple and cheap to produce. The projected use for the engine was to power a target drone aircraft. The engine they finished up with was essentially a turbo charger based device, not too dissimilar to the average amateur effort.

In the paper they discuss their arrangement for the air entry from the compressor into the combustor anullus. They state that the use of a tangential entry, creating swirl in the anullus, is to be avoided. However no reasoning is given for this.

Looking around, it seems that a lot of amateur engines use tangential entry, and I've even found places where it is encouraged.

So my question is, what are the reasons for either using or not using tangential entry to the anullus from the compressor?

[ausjet]

From my understanding you ideally want the air entry to surround the liner like many of the can type commercial small gas turbines do-Allison 250, ground apu engines etc. This diverging duct entry slows down the air, creates a more even pressurisation along the liner with less turbulence. To make a completely surrounded air entry on both sides of the liner is fairly difficult to do so we usually make the diverging duct entry on one side perpendicular to the combustor.

Tangential air entry to create swirl is the old way of doing it. Still works but not as efficient.

[finiteparts]

One of the primary challenges of using swirling flow in the combustor outer annulus is that swirling flows set up a radial pressure gradient with the lowest static pressure at the center of rotation.  The penetration of the liner jets and thus the mixing energy they provide is driven directly by the static pressure difference across the liner.  Thus, with the swirling flow, the lowest local static pressure occurs at the liner surface which is exactly opposite of what we want.   We want the highest local static pressure at the outer surface of the liner so that we have the highest capability to push the air into the combustor and provide the highest effective stirring energy. This is the primary reason why it is desired to have the flow in the annulus at a low velocity...since the local static pressure is highest at low speeds (i.e. static pressure is approximately equal to the total pressure).

In addition to providing the highest jet mixing energy for a given mass flow rate, lower annulus flow velocities also experience lower flow losses. Swirling flows suffer longer flow paths in the annulus, thus higher scrubbing losses and more chances for flow distortion or wakes from any element crossing the annulus. 

Finally, for a given mass flow through the axial cross-section, the swirling flow velocity will be higher than if it was totally axial.   Having a tangential velocity component causes the absolute velocity to increase in order to maintain the mass flux across the axial cross section.

You can make a swirl entry work, but in general, it is easier to get a non-swirling entry to work, especially for novices.

I hope that helps.  If anything is unclear, please let me know and we can discuss further.

Chris

[pitciblackscotland]

I have full docs on this subject, to large to post on here if someone would like a copy Can email a copy.

Cheers,
Mark.

[racket]

Hi Mark

Lotsa info in that one , I still have a look at the copy you sent me from time to time :-)

Cheers
John

[pitciblackscotland]

Hi John,
Yep some good reading in there. Wish i could get some info on the GTC20-1 Turbine i have.

Cheers,
Mark.

[racket]

Hi Mark

Now that one is a tad too rare :-(

Cheers
John

[britishrocket]

Thank you very much to Chris and to ausjet for the replies you gave. A concise, informative and detailed explanation. I'm beginning to see that obtaining an evenly distributed, low speed flow with the typical combustor designs used in amateur engines is not as simple as it appears. To my mind, the flow from the compressor needs to be arranged to enter the can of the combustor via a diffuser in such a manner that the flow runs parallel to the liner, and not perpendicular via a side entry as is the case with what I will refer to as the "accepted" amateur combustor design.

This "accepted" design would appear to have been arrived at more by expedience in terms of physically connecting the turbo compressors outlet to the can than any overriding consideration of how best to present the airflow. Of course, I can also see that a more elaborate pipework design may be required to achieve this parallel flow arrangement, not to mention a more complex head end design of the combustor. This more complex arrangement would be likely to incur flow losses and increased fabrication difficulty. I get that it is a trade off.

pitciblackscotland I'd really like to get a copy of the information you have, I will send you a message with my email address.

Thanks again to all. Any comments on what I've said above would be greatly appreciated. I'm learning all the time.

[racket]

Hi

I've been telling guys for years not to have tangential delivery tubes and not to have it positioned at the Primary Zone end , but unfortunately old habits keep them doing it , some tangentialy equiped engines simply haven't been able to run for the reasons Chris gave .

My TV84 engine from the 1990s had the delivery tube at the "bottom" ibb.co/kKVD4rk

So to the TV84 kart engine jetandturbineowners.proboards.com/attachment/download/46

And the Gt6041 engine jetandturbineowners.proboards.com/attachment/download/216

If you dump the air in "low" and from a "diffusing" funnel to spread the flow into the "plenum" surrounding the flametube at the scroll inlet , theres a better chance of having a more uniform airflow reaching the Primary Zone.

On smaller engines an axial "diffuser" up the side of the outer can is another option allowing air to be presented along the entire side of the flametube

Cheers
John

[britishrocket]

Hello Chris,

Thanks for that. Your photos are showing what I was thinking, that a good trade off interms of ease of construction and compressor outlet flow is to have the compressor feed the lower end of the can, before the tertiary zone holes.

[pitciblackscotland]

May 26, 2021 4:22:39 GMT -5 racket said:

Hi

I've been telling guys for years not to have tangential delivery tubes and not to have it positioned at the Primary Zone end , but unfortunately old habits keep them doing it , some tangentialy equiped engines simply haven't been able to run for the reasons Chris gave .

My TV84 engine from the 1990s had the delivery tube at the "bottom" ibb.co/kKVD4rk

So to the TV84 kart engine jetandturbineowners.proboards.com/attachment/download/46

And the Gt6041 engine jetandturbineowners.proboards.com/attachment/download/216

If you dump the air in "low" and from a "diffusing" funnel to spread the flow into the "plenum" surrounding the flametube at the scroll inlet , theres a better chance of having a more uniform airflow reaching the Primary Zone.

On smaller engines an axial "diffuser" up the side of the outer can is another option allowing air to be presented along the entire side of the flametube

Cheers
John

Hi John,
Yes the GTC20-1 built by Garrett turbine company has the delivery tube at the "bottom"

Cheers,
Mark.