Starting the math...
Using the AVL model I gave in a previous post, we'd like the wing loading for say a 6g pull-up maneuver. Say CLmax is 1.5 and load factor is 6 (using gross mass of 305 lb-mass), giving a necessary condition of 80 ft/s (54mph) and -21 deg elevator. At this test case, the trefftz plane shows:
This plot shows the loading really rolls off at about 12.5ft half-span. There are too many wing sections to easily look at the numbers, so let's drop down to 5 panels per wing half:
Surface # 1 wing
# Chordwise = 10 # Spanwise = 10 First strip = 1
Surface area = 179.309448 Ave. chord = 4.974011
CLsurf = 1.50090 Clsurf = 0.00000
CYsurf = 0.00000 Cmsurf = -0.01353
CDsurf = 0.50339 Cnsurf = 0.00000
CDisurf = 0.07366 CDvsurf = 0.42973
Forces referred to Ssurf, Cave about hinge axis thru LE
CLsurf = 1.32253 CDsurf = 0.44357
Deflect =
Strip Forces referred to Strip Area, Chord
j Yle Chord Area c cl ai cl_norm cl cd cdv cm_c/4 cm_LE C.P.x/c
1 -17.8428 4.2515 2.6609 1.5916 0.2041 0.4198 0.3729 0.0428 0.0047 -0.0402 -0.1256 0.359
2 -15.8461 5.0000 18.1066 5.0004 0.1971 1.1179 0.9994 0.1290 0.0453 -0.1015 -0.3534 0.352
3 -11.1640 5.0000 28.2017 6.7405 0.1283 1.4925 1.3546 0.4104 0.3418 -0.1260 -0.4765 0.343
4 -6.2763 5.0000 19.2959 7.3489 0.1029 1.6046 1.4793 0.5793 0.5221 -0.1328 -0.5206 0.340
5 -2.1627 5.0000 21.3896 7.6420 0.0951 1.6471 1.5386 0.6810 0.6267 -0.1366 -0.5431 0.339
6 2.1627 5.0000 21.3897 7.6419 0.0934 1.6471 1.5385 0.6810 0.6267 -0.1366 -0.5431 0.339
7 6.2763 5.0000 19.2959 7.3489 0.1029 1.6046 1.4793 0.5793 0.5221 -0.1328 -0.5206 0.340
8 11.1640 5.0000 28.2017 6.7405 0.1283 1.4925 1.3546 0.4104 0.3418 -0.1260 -0.4765 0.343
9 15.8461 5.0000 18.1066 5.0004 0.1971 1.1179 0.9994 0.1290 0.0453 -0.1015 -0.3534 0.352
10 17.8428 4.2515 2.6609 1.5916 0.2041 0.4198 0.3729 0.0428 0.0047 -0.0402 -0.1256 0.359
# Chordwise = 10 # Spanwise = 10 First strip = 1
Surface area = 179.309448 Ave. chord = 4.974011
CLsurf = 1.50090 Clsurf = 0.00000
CYsurf = 0.00000 Cmsurf = -0.01353
CDsurf = 0.50339 Cnsurf = 0.00000
CDisurf = 0.07366 CDvsurf = 0.42973
Forces referred to Ssurf, Cave about hinge axis thru LE
CLsurf = 1.32253 CDsurf = 0.44357
Deflect =
Strip Forces referred to Strip Area, Chord
j Yle Chord Area c cl ai cl_norm cl cd cdv cm_c/4 cm_LE C.P.x/c
1 -17.8428 4.2515 2.6609 1.5916 0.2041 0.4198 0.3729 0.0428 0.0047 -0.0402 -0.1256 0.359
2 -15.8461 5.0000 18.1066 5.0004 0.1971 1.1179 0.9994 0.1290 0.0453 -0.1015 -0.3534 0.352
3 -11.1640 5.0000 28.2017 6.7405 0.1283 1.4925 1.3546 0.4104 0.3418 -0.1260 -0.4765 0.343
4 -6.2763 5.0000 19.2959 7.3489 0.1029 1.6046 1.4793 0.5793 0.5221 -0.1328 -0.5206 0.340
5 -2.1627 5.0000 21.3896 7.6420 0.0951 1.6471 1.5386 0.6810 0.6267 -0.1366 -0.5431 0.339
6 2.1627 5.0000 21.3897 7.6419 0.0934 1.6471 1.5385 0.6810 0.6267 -0.1366 -0.5431 0.339
7 6.2763 5.0000 19.2959 7.3489 0.1029 1.6046 1.4793 0.5793 0.5221 -0.1328 -0.5206 0.340
8 11.1640 5.0000 28.2017 6.7405 0.1283 1.4925 1.3546 0.4104 0.3418 -0.1260 -0.4765 0.343
9 15.8461 5.0000 18.1066 5.0004 0.1971 1.1179 0.9994 0.1290 0.0453 -0.1015 -0.3534 0.352
10 17.8428 4.2515 2.6609 1.5916 0.2041 0.4198 0.3729 0.0428 0.0047 -0.0402 -0.1256 0.359
The column labeled cl_norm helps us get to the z-axis force at the local section. Backing out through the standard normalization CL=F_z/(Q * Sref) where Q = 0.5 * rho * V^2 ... section 1 (a tip) has the following loading:
Dynamic pressure:
q = 0.5 * rho * V^2 = 0.5 * 0.002378 slug/ft^3 * (80.7 ft/s)^2 = 7.743 slug/ft/s^2
Section force:
F_z = cl_norm * q * S_section = 0.4198 * 7.743 slug/ft/s^2 * 2.6609 ft^2
= 8.65 slug*ft/s^2 = 8.65 lb-force
(this is where I hate English units and love SI units ... kg makes more sense than lb vs lb-mass vs lb-force)
Carrying out these calculations on the other sections gives the following section forces:
section # | Yle (ft) | F_z (lb) |
1 | -17.84 | 8.65 |
2 | -15.85 | 156.74 |
3 | -11.16 | 325.93 |
4 | -6.28 | 239.75 |
5 | -2.16 | 272.80 |
6 | 2.16 | 272.81 |
7 | 6.28 | 239.75 |
8 | 11.16 | 325.93 |
9 | 15.85 | 156.74 |
10 | 17.84 | 8.65 |
The total sum of all F_z is 2007.74 lb, which is more than 6G's * 305lb, but reflects the additional loading due to dihedral angle (some aero force is pointing in the y-direction too). This tells that the outer two panels, from 18 to 17.68ft and 17.68 to 14.02ft combined have to carry approximately 165lb at the 6G loading case. Now the strut actually joins my wing at 139in (11.58ft), so we really should change the panel locations to correspond better with what is beyond the strut attach point. Here is a top-down view of the 10 panel wing for reference.
Going back to the 50 panel wing we started with at the top of the post, the wing outside the strut carries 269.5lb, the wing between the strut and the jury attach carries 353.8lb, and the inside wing between the jury attach and the root carries 378.5lb. Do note that these forces are not centered on the panels, especially the tip weight; rather, the lift distribution governs the location. That we can figure out from the lift distribution too ... but will wait for another day.
Here is the full 50 panel wing to get a better idea of the number of strips:
I drew up the spars in Solidworks and will run them through an FEA analysis to get an idea (*idea) of the stress distribution. Notably, there is area between the sleeves taken up by several wraps of electrical tape (as noted on Sandlin's drawings). The FEA will be assuming the walls touch and do not slip, which is not a conservative assumption. I'll chat with some folks at work to get some additional input.
Please, if you are a reader and note a mistake in my math or assumptions somewhere, please please let me know. I will not discard your input and would be happy to spend time working with you to get this right. Your life-saving thoughts are most appreciated :-)
2 comments:
I have been considering building one of these vehicles and have been searching for information about the structure. Voila! I finally stumbled over your work. I am not a trained engineer but I am somewhat familiar with what you are doing here with this analysis. Real live data. What a novel idea! Bravo. It looks to me like simply moving the lift strut location inboard an indeterminate amount would actually strengthen the wing structure in bending without gaining weight. Likewise, perhaps the (4) model with wire bracing might benefit by slight relocation of those fittings. There are some other, what I would consider non standard design elements in these aircraft that I would like to change. Yes, I know they seem to work OK, but there are a lot of eccentric joints in the wing that could be remedied very easily. And for the sake of a trailer many "pin" joints being retained with bungees that don't seem to me to be acceptable safety of flight items. At any rate I am fascinated with your work. Please keep it up. I'd like to start my glider soon and it looks like you're helping to solve some of my problems before I start cutting metal.
A couple additional comments. I was considering ordering the wing spar tubes in 16 foot lengths to avoid the stress concentrations at the splices. In addition the idea of wrapping electrical tape to fill the gap between the sleeves and the spar proper is not something I would tolerate. While it seems to be working (?) one of a couple things has to be happening in there where you can't see. The sleeve is being pulled off center and/or the rivets are not setting properly on the "gapped" side. Or, the spar itself is being deformed to conform to the smaller cross section of the sleeve. Electrical tape has no compressive strength nor tensile strength to deal with mismatches of this order. The solution would be to use a larger diameter sleeve and slice out a strip along the long axis to allow a force fit that would conform to the inner diameter of the spar. The Weedhopper Gypsy I built many moons ago used this technique and it worked very well.
Post a Comment