Logo
Builders' Plan Gallery  |  Hip Pocket Web Site  |  Contact Forum Admin (Account/Technical Issues)  |  Contact Global Moderator
December 06, 2021, 01:30:34 PM *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with email, password and session length
 
Home Help Search Login Register
Pages: [1]   Go Down
Print
Author Topic: EF-1 (NMPRA) pressure cowls  (Read 1014 times)
0 Members and 1 Guest are viewing this topic.
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« on: June 25, 2017, 10:41:15 PM »

Last year, elsewhere on the web, some of you might recall  I described how I installed a pressure cowl in my EF-1 (USA 4 cell Pylon Racers).  Based on inflight telemetry I was getting inconsistent cooling of the motor and windings. I think I’ve traced these variations to the gap between spinner to fuselage. As it is unfortunate that HIP doesn’t have a more active RC membership I feel it would be safe for me to post my speed secrets and latest solution to the EF-1 motor cooling problem, here on HIP. :0)

EF-1 is a highly spec’d entry level 4 cell pylon racing class in the USA. The aim is to keep the entry costs down and to emphasize pilot proficiency in pylon racing. And to that end it is very effective. As with most models there isn’t really much thought given to proper cooling. In racing this is a real concern as the motors are run at full power for the whole flight. Just to be clear EF-1 isn’t really all that demanding of the motor. But we really do want to cut down on cooling induced drag. The less air we need to use to cool the motor the less cooling drag there will be.

Just to give you an idea of how important cooling drag is, with piston powered aircraft going over 400 MPH, 3/4 of each additional horse power goes towards overcoming the additional cooling drag that the added horse power demands. This only leaves about a 1/4 hp to actually try to pull the plane faster.

While the EF-1’s cooling demands are not near as demanding, as these pylon racer fly at a modest 110 mph, but anything we can do to lower the drag will be of some benefit. Along with the reduction in drag, cooler motor windings perform better than hot winding as far as power generation in our electric motors. So with the drag reduction there is the promise of added power to the prop for any given amount of power extracted from the batteries.

Before I go any further I’d like to thank Horizon Hobbies for being one of the first big box hobby distributors to embrace this new class of racing. Most of the photos I’ll be showing are of the early AR-1 Pogo and latest Shoestring. Without Horizon Hobbies's (E-Flite) early enthusiastic support of this event I don't think EF-1 would be around today

The first thing one needs to do when racing is read the rule book. And as there is no mention as to cooling of the electrical components anything you can do to try to improve the design’s cooling efficiency is allowed. There is a prohibition on the use of battery heaters, that is the only thermal constraint I found in the EF-1 rules.

Most models as drawn and as manufactured only have a few cooling slots in the cheek cowls or under the spinner.  There is no provision to actually direct this air towards the motor. Historically most racers added some deflection baffles to direct air to the motors. This actually helped a lot. But much of the captured air still does not aid in the cooling of the motor.

The heat itself from the motor would actually deflects the cooling air away from the motor.
To solve this one needs to place the motor inside a tube and direct the air into the tube.

More to come.

All the best,
Konrad

Ratz Edit: Replaced content per writer's request
Attached files Thumbnail(s):
EF-1 (NMPRA) pressure cowls
« Last Edit: June 26, 2017, 06:02:23 PM by Ratz » Logged

Cut it twice and it's still too short!
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« Reply #1 on: June 26, 2017, 11:50:24 AM »

Here are some evolutionary photos of my progression in making an EF-1 pressure cowl.
At first I was going to just add baffles to the Pogo. These were going direct air across the whole rotating can of the motor.
The next day after a good nights sleep, I thought this was dumb and decided to add a tube and try to use the dead space behind the cowl as a plenum.
Attached files Thumbnail(s):
Re: EF-1 (NMPRA) pressure cowls
Re: EF-1 (NMPRA) pressure cowls
Re: EF-1 (NMPRA) pressure cowls
Re: EF-1 (NMPRA) pressure cowls
Re: EF-1 (NMPRA) pressure cowls
Logged

Cut it twice and it's still too short!
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« Reply #2 on: June 26, 2017, 01:55:06 PM »

Some more photos, showing the why's of adding a tube.
Attached files Thumbnail(s):
Re: EF-1 (NMPRA) pressure cowls
Re: EF-1 (NMPRA) pressure cowls
Logged

Cut it twice and it's still too short!
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« Reply #3 on: June 26, 2017, 05:56:30 PM »

Warning closely baffled motors actually suffer from the loss of radiant cooling, the baffles act as blankets. Do not run the motors without forward motion for any real length of time.

The baffles real are not helping much with the reduction in drag. They are helping with cooling, directing the high velocity air towards the motor. But most of this high velocity air is just blowing past and around the motor. It is not carrying away as much heat as it might otherwise be able to remove from the motor.
 
In the earlier posts I hinted at something with the term plenum and that we would try to cool the winding directly. To get the air to carry away more heat we need to slow it down. You might remember from your high school physics that if one slows down a stream of gas as the velocity slows the pressure rises.  It is this property that we will try to harness to improve our cooling systems efficiency. We do this by expanding the cooling air in a plenum. We take a metered amount of high velocity air from the air inlets and direct it into a large space. This large space is called a plenum. We will then try to push this air high pressure air down the motor.

Now there are several things working against us. Most EF-1 Pylon racers are using rear cross bracket mounting of the outrunner motor (the spec’s are calling out outrunner motors). This means that the prop is driven by the rotor end of the motor rather, than the stator mounting side of the motor. Due to the rotor end cap vents acting like a fan, the motor is actually trying to draw air from the rear (near the cross motor mount) towards the spinner. Yes, the motor wants the air to go the wrong way! The proper way to mount the motor would be with a nose ring mount, but that is beyond the scope of this thread.

The other issue is that whatever pressure we build up in the plenum would rather flow out the gap behind the spinner between the spinner and fuselage than do work and carry away heat. To make things worse the spinning spinner is actually trying to draw air from the plenum. So the motor rotor and spinner are actually working against us in trying to cool the motor.

The variability I've noticed in my temp reading I think can be traced to the slight variability in the fuselage to spinner gap as I change props. A gap much larger than 0.15mm makes a noticeable difference in the peak heat soak* reading I get at the end of the flight.

I choose not to take apart my motors and epoxy the temp probe next to the windings.  This make interpretation of the temp data a bit more complex. For any give amount of air and power I actually want to see a rise in the air temp at the point indicated in my drawing. This would indicate that the air is extracting more heat away from the motor. A rise in the temp reading might also mean that the motor was getting hotter for some unknown reason, like a leaf blocking an inlet. So what I'm using as my basis for determining the effectiveness of the cooling system is how hot does the motor get at the end of the flight when the plane is standing still. This is called the heat soaked condition. At speed the temp appears to be higher with the pressure cowl than without it. But the peak heat soak temp is less with flights that used the pressure cowl. And the difference between these two reading (at speed and heat soak) is less with the pressure cowl cooling system. So I think the pressure cowl works.

*You might have noticed that where I've placed my temp probe, that I'm actually measuring the temp of the cooling air at the discharge end of the tube. Normally I'd like to place the temp probe right against the motor windings. I place the temp probe where I did as the rules state no modification are allowed to the motor. Not wanting to get into a fight with the closed minded word smiths that tend to gravitate towards the competition classes, I placed the probe where I did.
« Last Edit: June 26, 2017, 06:11:32 PM by Konrad » Logged

Cut it twice and it's still too short!
TimWescott
Platinum Member
******

Kudos: 18
Offline Offline

United States United States

Posts: 1,237



Ignore
« Reply #4 on: June 26, 2017, 10:28:04 PM »

If you're going to go to the effort of making air move past the motor, making air move past the inside of the motor is probably a better bet.  I'm totally guessing here, but I'd mount the motor with the stator facing forward and with the plenum behind the prop sealed so that incoming air is forced through the inside of the motor.  A centrifugal fan on the back of the motor may or may not be necessary.  I'd even consider fairing the back of the spinner so that the air doesn't encounter any sudden turns as it enters the plenum.

I absolutely don't know if this would be a help or not.  It feels right, but aerodynamics is all about reality stomping on intuition.
Logged
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« Reply #5 on: June 26, 2017, 10:41:08 PM »

You're not wrong, that's effectively what the front nose ring mounting is all about. Using the rotor vents as the fan to draw air through the motor. The truth is that a quality motor has a rather high copper fill density and narrow air gap that there isn't enough air flow through the motor to draw away the enough heat to make much of a difference.

In racing you don't want to add a fan or anything else that takes away power from driving the prop.
Logged

Cut it twice and it's still too short!
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« Reply #6 on: June 27, 2017, 09:29:01 AM »

With that ground work laid, here is the HIP exclusive! The 2017 improvement is a change from a face seal (the gap between the fuselage and spinner) to a labyrinth seal (in blue)!
Attached files Thumbnail(s):
Re: EF-1 (NMPRA) pressure cowls
Logged

Cut it twice and it's still too short!
Starduster
Platinum Member
******

Kudos: 39
Offline Offline

United States United States

Posts: 1,470




Ignore
« Reply #7 on: June 27, 2017, 11:04:34 AM »

Probably a silly question, but have you experimented with a standard or slightly modified NACA Cowl?
Logged

"We have done the impossible, and that makes us mighty..."
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« Reply #8 on: June 27, 2017, 12:00:19 PM »

That is basically what a pressure cowl is. Take a little  high velocity air slow it down by expansion ahead of the heat source. And then accelerate it with the heat source. The NACA Cowl is a special design for radial engines. As most EF-1 are modeled after the Reno F-1 sport class racers that have flat engines. The accepted term is pressure cowl.

I'm just letting the air discharge into the fuse to be used to cool the ESC and batteries.  Now I could duct the exhaust to produce some thrust by the Meredith effect. https://en.wikipedia.org/wiki/Meredith_effect. But this is  highly unlikely at this scale. I would then still need to add some cooling drag the cool the ESC and batteries.  And the added weight for the discharge ducts has a great penalty at the very high coefficients of lift (G's) we see in the hard sharp pylon turns.
« Last Edit: June 27, 2017, 12:37:14 PM by Konrad » Logged

Cut it twice and it's still too short!
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« Reply #9 on: June 27, 2017, 12:22:35 PM »

A labyrinth seal works by successively dropping pressure across the knife edges so that at the end there is very little flow. The resut is that it is easier for the air to go down the tube than leak out past the spinner.

To make the knife edges I’ve chosen to use the thin plasticized paper often found in transit passes. I then made the spacers from 1.5 mm balsa with a 21mm hole punched in it. These spacers are effectively mini expansion chambers. And for strength the end faces are 0.7mm plywood. My spinners have a smooth 15mm boss. This boss is what the knife edges will be working against. For a labyrinth seal to work the clearance between the knife edges and the boss need to be small , in the order of 0.1mm to 0.15mm. This clearance is just as critical as the fuselage to spinner gap with the face seal, but it should be easier to maintain.

To minimize the inevitable material pick up as the boss touches the knife edges I grit blast the spinner boss.
Attached files Thumbnail(s):
Re: EF-1 (NMPRA) pressure cowls
Re: EF-1 (NMPRA) pressure cowls
Re: EF-1 (NMPRA) pressure cowls
Re: EF-1 (NMPRA) pressure cowls
Logged

Cut it twice and it's still too short!
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« Reply #10 on: June 27, 2017, 03:53:55 PM »

As the tube and seal are set up based on the motor center line, now would be a good time to talk about airframe set up. You really won’t have much room to adjust the thrust line after you glue in the cooling tube and labyrinth seal. The tube has 1.5 mm to 2 mm clearance with the motor and the labyrinth seal has 0.1mm clearance with the spinner boss! So test fly and make your adjustments before making the pressure cowl.

Here are some of my set up values. I’ve found that all the EF-1 I have set up needed about 5mm to 6mm of right thrust as measured by the difference in the 8 inch prop tips from the rudder post.  On the 3 E-Flite Shoestrings  I’ve had to make the left motor stand-offs 1mm longer than the right stand-offs. Also as the upper left stand off would hit the left part of the cowl, I’ve needed to sand away about 2 mm from the upper left stand-off and adjacent firewall (see attached photo).

As there is no clearance for the cowl to move you will want the cowl to key (index) solidly to the firewall.  I do this by filling any space between the firewall and the cowl with a thick mixture of 5 minute epoxy and micro balloons. As the mixture is starting to set on the firewall I cover it with cling wrap, to keep the epoxy from bonding to the cowl, and quickly shove the cowl in place. Once the epoxy sets I remove the cowl do any clean up and repeat to fill in any voids or add more support if I feel it is needed. This has taken on average about 3 attempts to get a good copy on the inside of the cowl cross section imprinted on the firewall.

On the Great Planes Proud Bird I’ve only had to add about a 0.2mm shim to the left motor stand offs. On the one Pogo I’ve set up I had to add 0.5mm of shims to add down thrust. Once you have the thrust line you want tape up the motor and spinner boss to take up the clearance and glue the cooling tube and labyrinth seal to the cowl.

Once everything is cured you are going to want to use the back of the spinner and some sandpaper to sand the spinner ring square to the motor shaft. While I’m confident that the labyrinth seal will work I also like the idea of trying to keep the face seal. Besides the small gap looks cool!
Attached files Thumbnail(s):
Re: EF-1 (NMPRA) pressure cowls
Logged

Cut it twice and it's still too short!
Konrad
Titanium Member
*******

Kudos: 39
Online Online

United States United States

Posts: 2,314


Topic starter
Measure twice cut once



Ignore
« Reply #11 on: June 28, 2017, 09:22:26 AM »

Now that I’m coming to the end of discussion it is time for some clean up and disclaimers. I size my motor tubes for a diameter that is 3mm to 4mm larger than the motor diameter. My pressure cowls add about 14 to 21 grams of weight. This depends mainly on the weight of the paper motor tube. I have not data on the engine’s temp for any of the stock (non-baffled set up) be them ether Great Planes of E-Flite models. I’m able to block off 50% of the cooling air inlet and still not see any rise in motor temps.

Now the disclaimer: I have not seen any reduction in my lap times with these pressure cowls. It is hoped that with the cooling running motor there is more power to the prop. And that with the ability to use smaller inlets that there is less drag. These improvements should help with speed. But since pylon racing is about flying the course, speed really is not the main criteria pilot proficiency is. These pressure cowls do nothing to help one fly better.

All the best,
Konrad
Logged

Cut it twice and it's still too short!
Pages: [1]   Go Up
Print
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Valid XHTML 1.0! Valid CSS!