marvnero
Member
Joined: February 2017
Posts: 33
|
Post by marvnero on Mar 17, 2017 10:34:34 GMT -5
|
|
|
Post by racket on Mar 17, 2017 17:02:53 GMT -5
Hi Marvin
There seems to be a lot of hole area in those two walls , more than expected for a modest sized compressor inducer , could you provide your calculations please.
Also if this flametube is to feed into a turbo scroll how are you going to arrange the outflow from the flametube ?
Cheers John
|
|
marvnero
Member
Joined: February 2017
Posts: 33
|
Post by marvnero on Mar 23, 2017 18:01:18 GMT -5
Hi John, I checked my calculations and found out that I made a huge mistake right at the beginning...I somehow calculated a way too high temperature rise across the compressor. As I had to calculate everything again anyway, I tried to use the Kamps chamber's properties even more to get the best possible result. I began by calculating the temperature rise across the compressor at a pressure ratio of 3.5 with an efficiency of 77% which turned out to be T_2= 485°K at 20°C ambient temperature. The air density at p_2 and T_2 is 2.5 kg/m^3. These are my reference points to estimate a hole area based on the Kamps design. I want to keep the same airspeed entering the chamber. I calculated the density of the Kamps turbine at 1.85 kg/m^3. I added up all the holes of the kamps chamber giving me a total area of 18.74 cm^2. With these data I calculated the Kamps an mine volume flow. To get the required hole area for my chamber I used the equation velocity=const.= V_k/A_k= V_m/A_m.--> A_m (my flow area) = 52cm^2. As the stick flow area of the Kamps chamber is approx. 10% of the total flow area, my stick flow area has to be 5.2cm^2. To get the needed areas for the primary and secondary zone, I took the first 25mm of the kamps chamber as the primary zone and the rest as the secondary zone. Using the hole areas of these zones and multiplying them with a factor of A_totalMine/A_totalKamps= 2.774, I figured out the required hole area for each zone of my combustion chamber. Your suspicion that my first design has too much flow area was totally right! Thanks for telling me this, otherwise I probably would've wasted days building something that would never work xD In the end that gave me as smaller total area, smaller stick area but as a result a slightly bigger primary zone area. Is the even the right way to get a working combustion chamber? EDIT: To feed the turboscroll I came up with the design shown in the attached pic. Do you think that it'll work? postimg.org/image/kxz4re56h/Cheers, Marvin
|
|
|
Post by racket on Mar 23, 2017 18:54:38 GMT -5
Hi Marvin
Scaling directly from the Kamps engine has limitations due to its modest pressure ratio and T I T temperatures.
I generalise and simply measure the cross sectional area of the whole inducer ( hub included) , so for your 67 mm inducer its 3525 sq mms , 10% for the sticks ID so 352 sq mms , their OD heating surface needs to be 6 times inducer area or 21,150 sq mms, total Primary Zone area is 30% as we can run higher turb temps ( more fuel and air) with a turbo turbine made from Inco , so if 10% sticks that leaves 20% for Primary holes or 705 sq mms , I split that ~60/40 between outer and inner walls as theres a tad more room on the outer wall for holes and theres less room needed for the flow between shaft tunnel and inner wall which can be restrictive .
20% or 705 sq mms for Secondary holes , again split 60/40 out/inner.
50% or 1762 sq mms for the Tertiary holes , 60/40 , though this can be "adjusted" a bit depending on the need for making it "look right" .
Cross sectional area of the flametube needs to be at least 3 times inducer , so >10,575 sq mms.
Yep , your drawing of the flametube should be OK , though how and where are you mounting the evap sticks , you'll be needing J sticks rather than straight ones.
Cheers John
|
|
marvnero
Member
Joined: February 2017
Posts: 33
|
Post by marvnero on Mar 24, 2017 18:29:48 GMT -5
Hi John, Thanks for checking my calculations! I should've asked you how to determine the hole area before my quite time consuming attempt to scale up from the Kamps chamber...next time I'll ask right at the beginning I revised my hole area with the values you recommended. I'd really like to avoid using J sticks as this would completely mess up my hole placement and by doing so I'd have to CAD every part again. Instead I'd like to use a design quite similar to the Kamps design shown in the attached pic. PS: I forgot to place the cone at the end of the inner flametube. The sticks look poorly but this is just a qick and dirty mock-up to get your opinion of the design. What do you think of this evaporator design? postimg.org/image/64zlh1l3d/Cheers, Marvin
|
|
|
Post by racket on Mar 24, 2017 19:09:38 GMT -5
Hi Marvin
Yep , thats the same sort of design I ended up with , you may need to have the sticks angled a tad so that the outlet is centralised in the annulus .
Cheers John
|
|
marvnero
Member
Joined: February 2017
Posts: 33
|
Post by marvnero on Apr 3, 2017 17:40:31 GMT -5
Hi John, Alright, I'll focus on that stick design. A problem I neglected quite some time was how to feed the air into the outer casing of the combustion chamber. I had some rather complex designs in mind which woud have made my annular combustor not much smaller than a tube-style combustor... To keep the advantage of the more compact packaging using an annular combustor, I kept looking for a suitable design. My hairdryer was an inspiration for my latest and most compact design. What do you think of this design? The cross sectional area at the combustors entry is 6600 sqmms. postimg.org/image/fgtlbg29h/Cheers, Marvin
|
|
|
Post by racket on Apr 3, 2017 19:44:56 GMT -5
Hi Marvin
Not a good design , you're creating a vortex , and for air to get to the centre the velocity needs to increase and the static pressure to drop , exactly the reverse of your compressor diffuser scroll.
Cheers John
|
|
marvnero
Member
Joined: February 2017
Posts: 33
|
Post by marvnero on Apr 20, 2017 17:57:59 GMT -5
Hi John, Please excuse my late reply, but there was not a lot time to spare for thinking about an improved design during the last weeks. I haven't thought that creating a vortex would be that bad as Tubular combustor designs lead to a swirl as well ( at least as far as I know). I went back to my initial idea of a longitudinal air flow design as shown in the attached image. In the drawing the air is supposed to flow from left to right. The small cone on the left side (entry) should keep the air entering the combustion chamber from directly flowing into the inner flametube at a high velocity and thereby creating a higher pressure than on the outside of the outer wall while decelerating in the inner tube. Is this design more likely to work properly in your opinion? Link to my latest drawing: postimg.org/image/i6fg4hk0z/I'm not quite sure how thick the outer wall of the C-chamber should be. The max. pressure ratio should be around 3.2 creating a pressure difference of 2.2 bar between the inside and the outside of the outer wall. How thick should the outer wall be in your opinion? ( I'd like to use "regular" quality steel for the outer wall as it is less complicated to work with) Cheers, Marvin
|
|
|
Post by racket on Apr 20, 2017 18:42:46 GMT -5
Hi Marvin
The airflow should be below 100 m/s by the time it enters the combustor can , the "cone " will stop any slight pressure difference , but make sure you have plenty of flow clearance between it and the endwall of the FT so that there isn't any restriction to flow .
Swirl inside the flametube is beneficial for combustion , but it has to be the right kind of swirl in the right place , having all of the delivery air swirling is a recipe for problems .
You should be able to use <2 mm thick steel sheet for the can , but it will depend on the final OD of the can, I used 1.6 mm stainless for my 300 mm dia 10/98 engine can .
Cheers John
|
|
marvnero
Member
Joined: February 2017
Posts: 33
|
Post by marvnero on Oct 6, 2017 16:49:57 GMT -5
It's been a while since my last post but finally I got all my Bachelor exams done and found some time to continue building my combustor. I built an Anders's style sheet metal bending device and I'm absolutely stoked about how good the results of using that device are! I got most of the parts cut out and ready to bend/ weld. Attached are pictures of the probably most difficult to bend part of the combustor. Huge thanks to Anders for his post about that sheet metal bending device! Without that post I'd probably still be hammering sheet metal around pipes for countless hours I have to build a radial "compressor" for our workshop as my extensive cutting and welding causes it to fume up even with our ventilation system turned on... But I'll try to continue building the combustor simultaneously and post updates frequently. john sorry for the late reply... I'll be using 1.5mm steel for the can as I'll be running relatively low P2s and the can is only to test the flametube. Once I got that figured out, I'd like to build a more compact design with the flametube around the shaft tunnel just as you and Anders did. Links to the photos, postimg.org/gallery/2npkqxmm4/PS: Please excuse the poor welding at the bottom of the part.I never welded stainless that thin before and used waaay to much current--> big fat hole I had to fill in... Cheers, Marvin
|
|
marvnero
Member
Joined: February 2017
Posts: 33
|
Post by marvnero on Nov 23, 2017 17:57:23 GMT -5
Hi guys, here's another update of my build. I finally got all the holes drilled into the Combustors walls...man that killed a handful of drills The Anders style sheet metal rolling device came in handy again and the result is not as bad as may look on the pictures. My welding is still not on point, but I'm getting definitely more confident welding that super thin sheet metal. Attached you'll find some pics of the part I finished today. I'll keep you updated about further progress. Edit: attaching pictures didn't work... You have to followthe link sorry :/ postimg.org/gallery/18p43k3oo/Cheers, Marvin
|
|
marvnero
Member
Joined: February 2017
Posts: 33
|
Post by marvnero on Nov 28, 2017 17:38:19 GMT -5
Hi folks, Here`s another update on my build. A little bit unconventional this time. We have a small Workshop in my dorm and we recently bought a Chinese 3040 CNC machine (of course that was my stupid idea ). After I spent countless hours on setting it's software up properly and even longer generating useful G-codes with Fusion360, it's time to properly test it! Until now I cut just easy geometries like triangles and it's doing quite good there but that's not a proper task for a CNC machine. I created a CAD file for 2molds to bend the front face of my annular combustor in a donut-like shape. That was done in about 10 minutes but creating a CAM file was a pain in the ass... I hope I'm getting faster over time... Unfortunately it's too late to test it today, but I'll keep you updated about how the parts turned out! Attached you'll find a screenshot of all the confusing toolpaths...I hope no one who's actually a CNC programmer sees that, it's very uneficcient for sure Link to Image: postimg.org/image/wzsaq0vl3/
|
|
marvnero
Member
Joined: February 2017
Posts: 33
|
Post by marvnero on Dec 16, 2017 15:03:10 GMT -5
Hi guys, Here's yet another update on my build. I already milled one half of the mold and it turned out quite good. It would've been close to perfect if I got the CAM code right. Next time it'll look better for sure. That cheap chinese CNC mill seems to be alright for some backyard engineering I also finished the inner combustor liner. I tried my best to get a good looking weld this time, but as you can see on the attached images it's still far from perfect...I'm wondering about why the weld is completely black- maybe too much heat? BTW do you seal the flametube and flood it with argon while welding it as one should do when welding regular pipes ? Links to the images: postimg.org/gallery/y43klvl8/Cheers, Marvin
|
|
|
Post by racket on Dec 16, 2017 15:39:50 GMT -5
Hi Marvin
Your first pic of your flametube has me confused , it doesn't look "right" .
Is the cone at the Primary end or the outlet end , the size of the wall holes sorta indicate its at the outlet end but the actual outlet hole is pretty small :-(
Cheers John
|
|