# Ultralight Carbon Spar...

Discussion in 'Composites' started by Senna, Jan 31, 2003.

### Help Support HomeBuiltAirplanes Forum by donating:

1. Jan 31, 2003

### Senna

#### Well-Known Member

Joined:
Jan 17, 2003
Messages:
55
0
Location:
Crooked River Ranch, Oregon
After much thought and number crunching I've decided to design and build a carbon composite spar for an ultralight I've had under construction for some time. The original spar design called for sitka spruce/ply and was engineered for a +3.5 g load rating. I actually constructed the spars as designed and have the center section complete with nose ribs, sheeting, main spar, mid ribs, rear spar, trailing edge ribs, and trailing edge. Unfortunately the outer spars were damaged during one of my many moves during this long building process.

I became enamored with this ultralight design way back in the late '80's after seeing it in an issue of Sport Aviation (I don't recall the issue). Actually in the photograph there were two of this particular ultralight parked on the grass side by side. Pod fuselage with boom, V-tail, tricycle gear, cantilevered wing, & pusher engined. I ordered and received plan set #340.

Over the years I have worked in fits and starts on this ultralight. I'm about to begin again. With the advancement of technology and materials... carbon fiber has become not only attractive but affordable. A recent discovery (a discovery for me anyway... as this product has been available for several years) has renewed my interest in completing and flying my plane. I found a mention of a carbon fiber pultruded rod called Graphlite™ during one of my many sessions surfing the Internet. I came across this material while viewing and reading of the construction of the Wes Whitehurst CARBON MONARCH on Jim Marske's web site. Marske has a link at the bottom right on his home page that took me to a paper he had written extolloing this Graphlite™ rod. If interested... you can read it here...

http://www.continuo.com/marske/carbon/carbon.htm

I first thought to incorporate the carbon rod into the original spruce spar by epoxying several of them in saw kerfs cut into the upper and lower spar caps. Once constructed I intended to proof this modification by statically load testing the spar. I began thinking about how I would conduct this necessary proofing... the rig that would be required... the method of loading the spar... the amount of weight for specific g loadings. It was while working with the numbers that I came to the conclusion that the original design, while benefiting from the incorporation of the carbon rods, would not take enough advantage of the tensile and compressive qualities the carbon possesses. I decided to completely scrap the idea of a spruce/carbon modified spar.

So... I'm embarking on a new adventure. One of exploration of an exciting product that has amazing structural qualities. My intent is now to engineer, design & construct a carbon main spar for my ultralight.

A spar mold for layup and vacuum bagging will be the first item of order. Since the original spar dimensions will be used... I'll have need of a 12 foot X 9-1/2" X 3/4" form to layup the new outer panel spars. The following drawing is the approach I plan to take.

More as it happens...

Senna

#### Attached Files:

• ###### sparmold.jpg
File size:
74 KB
Views:
2,419
Last edited: Jan 31, 2003
2. Jan 31, 2003

### Jman

#### Site Developer

Joined:
Oct 22, 2002
Messages:
1,881
58
Location:
Pacific NW, USA!
Senna,

MDF is a great choice for the mold in my opinion. I've worked with that stuff quite a bit and it is extremely flat and easy to work with. It doesn’t droop very much when placed on a span between supports either. I'm excited to hear how this project unfolds.

Jake

3. Jan 31, 2003

### Senna

#### Well-Known Member

Joined:
Jan 17, 2003
Messages:
55
0
Location:
Crooked River Ranch, Oregon
Hi Jake,

The fact that MDF is so flat makes it a logical choice. I still need to build the work benches and have resorted to making my own truss members for strength and straightness. 7/16" OSB web with 1-1/2" fir caps. I hand plane to true on the upper cap. It's inexpensive which is important too! I'm making two 6'-8" benches as I'm using hollow core doors atop the homemade truss as the top surface. Square...plumb...& level...!!! Ayup!!!

I've figured out the carbon material required for the new center section main spar and both outer panel main spars. I will be using 5.7oz Twill 3K 50" wide carbon fabric. Two layers of the 5.7oz cloth laid on a 45º bias will encapsulate the .092" X .220" rectangular Graphlite™ pultruded carbon rods needed for the caps. The 6 foot center section will use three stacked rods top and bottom... full length. The outer spars will use three rods in both the upper and lower caps. One rod will run the entire 12 foot length of the spar... the 2nd rod will stack atop the 1st and run out 9 feet. The 3rd will stack atop the first two and run out 5 feet.

I will need 6 yards of the 5.7oz Twill fabric and 140 feet of the Graphlite™ rod. That comes out to $141.00 for the cloth and$123.20 for the pultruded rod. That is less than the cost of the aircraft sitka spruce and aircraft ply to build the original spruce and ply design. Haven't researched epoxy resins as yet...

Here's the cut schedule for the 6yds of carbon cloth needed... not certain if anyone is interested... but it is colorful!!!

Senna

#### Attached Files:

• ###### carboncut.jpg
File size:
66.7 KB
Views:
1,677
Last edited: Jan 31, 2003
4. Feb 1, 2003

### George Sychrovsky

#### Banned

Joined:
Jan 25, 2003
Messages:
430
27
Location:
Shirley airport MA
are you sure thats enough carbon rods ? I made a preliminary estimate on my design which has a span 17 ft gross weigt 550lb spar height 6 inches and I figured I would need about 6 rods in the the upper cap

5. Feb 2, 2003

### Senna

#### Well-Known Member

Joined:
Jan 17, 2003
Messages:
55
0
Location:
Crooked River Ranch, Oregon
George,

Just ran some numbers and for my 410lb gross @ +5 g loading... I come out with max limit bending moments of:

0% span >>> 84,839 in-lbs
10% span >>> 67,871 in-lbs
20% span >>> 51,752 in-lbs
30% span >>> 37,329 in-lbs
40% span >>> 25,452 in-lbs
50% span >>> 16,968 in-lbs
60% span >>> 10,181 in-lbs
70% span >>> 5,939 in-lbs
80% span >>> 2,545 in-lbs
90% span >>> 1,103 in-lbs

With my spar being symmetrical in design I feel I'm pretty well covered structurally in both positive and negative loads with the three rods per cap. The carbon/carbon spar (carbon spar caps and carbon shear web) will have carbon flanged channel stiffeners epoxied to the shear web at all mid-rib locations. These will help strengthen the shear web and flush up the spar rear face to the spar flange for ease of rib attachment. I call them rib hats.

Senna

File size:
39.2 KB
Views:
3,297
6. Feb 2, 2003

### Senna

#### Well-Known Member

Joined:
Jan 17, 2003
Messages:
55
0
Location:
Crooked River Ranch, Oregon
After playing with lots more numbers while watching the 24 hours of Daytona... I've changed the number of rods in the spar caps... adding one additional rod top and bottom. I had the spar originally designed with a spanwise diminishing 3 rod stack per cap. The problem isn’t one of tensile or compressive strength in the 3 rod stack... it was a problem with deflection! Because the carbon rod material is so strong and so little rod is required the anticipated beam deflection was problematic and an additional rod was added as a means of stiffening the caps. This only added $29.92 to the total cost for the rod. I then went to a proven design and ran the numbers for that against my design. Although it's an anecdotal comparision... the results are very interesting! The Marske Monarch flying wing sailplane has a wingspan of 42.6 feet and uses 112 feet of Graphlite™ rod per wing in a spanwise diminishing 4 rod stack. Given that the Graphlite™ .092 X .220 rectangular carbon rod weighs 0.2053571 ounces per linear foot... this amounts to 22.999995 oz of carbon rod in each 21.3 ft spar. My spar will now use 63 feet of Graphlite™ rod per wing in a spanwise diminishing 4 rod stack. This amounts to 12.937497 oz of carbon rod per each 12 ft spar. My 12 ft spar happens to be 56.33803% of the Monarch’s 21.3 ft spar length. If you multiply the Monarch’s total carbon rod usage of 112 ft by 56.33803%... you end up with 63.098593 feet. Remember I’m using 63 feet of rod per spar! If you multiply the Monarch’s total carbon rod weight of 22.999995 oz per spar by 56.33803%... you come out with 12.957744 ounces. My spar comes out using 12.937497 oz of carbon rod per spar which is only .020247 ounces less material. The ratio of footage and ounces in my spar is almost identical to that of the Monarch’s!!! Like I said it's anecdotal... but it sure makes me feel better knowing all my amateur number crunching equates to a 'real world' application. Now I need to build the workbenches and spar mold. More as it happens... Senna Last edited: Feb 3, 2003 7. Feb 2, 2003 ### George Sychrovsky ### George Sychrovsky #### Banned Joined: Jan 25, 2003 Messages: 430 Likes Received: 27 Location: Shirley airport MA You are really good with them drawings, My hat off. 8. Feb 3, 2003 ### Senna ### Senna #### Well-Known Member Joined: Jan 17, 2003 Messages: 55 Likes Received: 0 Location: Crooked River Ranch, Oregon With the change of an additional rod to the spar cap rod stacks... it has caused the inevitable cascade of adjustments required to those items downstream. The redesign now uses a four rod pack of diminishing rod length for each spar cap. The drawing below illustrates the intended carbon rod pack spar cap layout. Senna #### Attached Files: • ###### rodstack.jpg File size: 88 KB Views: 2,275 9. Feb 4, 2003 ### Senna ### Senna #### Well-Known Member Joined: Jan 17, 2003 Messages: 55 Likes Received: 0 Location: Crooked River Ranch, Oregon First order item... The workbench is the first thing needed for this project. Unfortunately my last move forced me to abandon my rather large and heavy benches. So... new ones are in order. To get the bench tops as flat as posible I've begun building homemade truss sections. That's nothing more than a crude spar with upper and lower caps seperated by a shearweb. The local Home Depot's selection of lumber leaves a lot to be desired. By milling my own cap lumber I'm able to achieve impressive results. Strong and straight! The following pic is not all that impressive... but it's a start! Senna #### Attached Files: • ###### homemadetruss.jpg File size: 58.9 KB Views: 1,840 10. Feb 19, 2003 ### Senna ### Senna #### Well-Known Member Joined: Jan 17, 2003 Messages: 55 Likes Received: 0 Location: Crooked River Ranch, Oregon I've been trading e-mails with Jim Marske with regard to the carbon spar design. Jim was kind enough to go over my numbers and confirm the adequacy of my rod pack and spanwise rod lengths. Based on the following... at 410lbs gross and +5g loading... my wing root_will see 84,839 in-lbs of force. The spar depth is 9.5 in. Dividing the load force of 84,839 by the 9.5 in. spar depth we get_ 8,930.42 lbs. Divide that result into a conservative derated rod strength of 200,000 psi (tested to 320,000 - 330,000 psi) we arrive at .044652 sq in. This is the needed cross section of carbon rod. Divide that result by a single rod cross section of .02024 sq in (.092 X .220 rectangular rod) and that equals 2.21 rods. (3 rods = 6.6g's) The following is Jim Marske's reply... Looks like you are well in the ballpark and your rod lengths seem OK._ I do not know the spar depth at these stations but you can fill in the blanks._ I tend to drop off the rods at equal spacing from root to tip._ This gives the spar more of a true arc in the deflection mode although I get some wing skin buckling at the root first._ However, less bending in the aileron area means the ailerons are less likely to bind under load. The amazing thing is that the 108.4 ft of rod (based on your design lengths) weighs only 1.41 lbs for both outer panels and at$.88 / ft = \$95.00

With the spar cap design taken care of... now I'm having some difficulty with the shear web torsional strength and need to work on that aspect of the design a bit more. It's starting to get slightly more complicated than I originally thought. If I was starting from a clean sheet of paper I wouldn't be hitting as many snags. Attempting to reverse engineer a carbon design into the existing original dimensions has gotten tricky. Some things may need changing.

Senna

Last edited: Feb 20, 2003
11. Feb 21, 2003

### Senna

#### Well-Known Member

Joined:
Jan 17, 2003
Messages:
55
0
Location:
Crooked River Ranch, Oregon
I'm running into difficulty attempting to create the final root spar width of 3/4 inches so that the root attach fittings mate to the center section wing attachment fittings. The laminated fiberglass fitting attach blocks need to be securely fixed so that the rod pack translates the load into the fitting blocks and the fittings themselves. I can't figure out an elegant way to make the transition from the .120 shear web flange thickness to the needed 3/4" without it becomming a real kludge.

I've also read that an all carbon structure is virtually impossible to inspect and detect damage. It just doesn't show itself as such and gets overlooked. One fix is the use of a lightweight fiberglass cloth laminated to the exterior of the carbon structure that shows readily any compression, elongation, or impact damage. More work... more money...

This has just gotten too complicated for this particular application. I've decided to go back to the idea of bonding the Graphlite carbon rods into a dado cut into the upper and lower spruce spar caps of the original spar design. It won't utilize fully the carbon rods tremendous strength... but the spruce will benefit greatly with the addition.

Senna

Last edited: Mar 6, 2003
12. Mar 6, 2003

### dannicoson

#### Member

Joined:
Feb 27, 2003
Messages:
15
0
Location:
Round Rock, TX
Don't give up so easily! You are almost there!

At the root of your spar where you want to have the fiberglass bolt plates, instead of trying to glue a finished block in place, build up 5-6 thick layers of glass at a time starting a foot or so out from the root end. Let the final thickness of 3/4" be there for the last 6-8 inches.

I use a tri-axial glass fabric from Aircraft Spruce, their part number KYNTEX DB-170, to build up my bolt pads because it is very thick. Make sure you compress these buildups during resin cure to keep them very dense in fiber content.

By building up with a tapered stack of glass you will more gradually feed the loads from your shear web and spar cap material into the bolt pads. I used a bearing strength of 50,000 psi on my bolt pads for my own tests. I made sure I had a significant margin of safety for the bolts as they went through the bolt pads.

This method will ad some weight right at the root of your spar but you are saving SO much weight by using the carbon spar to begin with. Going back to spruce gives it all away. As far as inspecting over the years, how do you know your spruce isn't rotting from within? With the composite spar you can do a load test of your wing every so many years. As long as you don't see any permanent deflection you're still good-to-go.

You mentioned trying to figure out the torque exerted on the spar from airfoil and aileron forces. These torque loads are magnitudes lower than the bending forces you have been looking at. The outer skins of your wing will carry all these loads up to the root. The wing root rib (a simple foam sandwhich should carry that load to the 3/4" thick root we just built into your spar above. From there the twisting load goes through your metal bolt plates to the spar center section and right into the fuselage side structure. Also, if you have a rear spar attachment on this wing it will unload most of this load too.

Check out "Stress Without Tears" to see how Tom Rhodes addresses a lot of these loads. He keeps it pretty simple and straight forward.

Happy designing!

Dan Nicoson

13. Mar 6, 2003

### Senna

#### Well-Known Member

Joined:
Jan 17, 2003
Messages:
55
0
Location:
Crooked River Ranch, Oregon
Dan,

Thanks for the encouragement! Your description of a tapered glass build up at the spar attachment pads was exactly what I had envisioned and what I had termed the 'KLUDGE' in my last post. The weight penalty was anticipated to be unacceptable.

The analysis paralysis I'm experiencing has been over my needing to create a final spar thickness of 3/4" so that the attach plates mate with the required interference fit to the center section fittings. Also I have the mid-ribs and nose ribs all pre-cut and the airfoil cross section requires a spar thickness of 3/4". In other words I'm attempting to shoehorn this design into inflexible parameters.

I sure appreciate your thoughts on the matter though. Ruminating material!!!

Thanks for taking the time...
Senna

14. Mar 7, 2003

### Senna

#### Well-Known Member

Joined:
Jan 17, 2003
Messages:
55
0
Location:
Crooked River Ranch, Oregon
Dan,

The following drawing is what I had in mind and what I called the KLUDGE. It parallels closely your suggestions. It just appeared to me to be an inelegant solution to the design problem.

Senna

#### Attached Files:

• ###### rootbuildup.jpg
File size:
45.8 KB
Views:
1,846
15. Mar 19, 2003

### dannicoson

#### Member

Joined:
Feb 27, 2003
Messages:
15
0
Location:
Round Rock, TX
Senna,

What drawing tool are you using to illistrate your spar? That looks cool!

As far as this being an "inelegant solution", I think it is far more elegant than going completely back to wood. You've already calculated a huge weight savings and now you give it up for 2-3 pounds of extra material?

The reality is that you are concentrating the stress in the wing by having a spar connection. You can't avoid that except if you have a full length spar. Even then you will have to have some "doublers" where you connect the spar to the fuselage. Anytime you concentrate forces you have to have the right amount of material to accomodate the force.

Have fun,

Dan

16. Mar 19, 2003

### Senna

#### Well-Known Member

Joined:
Jan 17, 2003
Messages:
55
0
Location:
Crooked River Ranch, Oregon
Drawing program...

Dan,

I'm an Apple Macintosh user and do all my drawing with a program called Freehand. One semester of Jr High School mechanical drawing has also served me well. I appreciate the encouragement and the somewhat balmy rebuke!

Senna

17. Mar 19, 2003

### dannicoson

#### Member

Joined:
Feb 27, 2003
Messages:
15
0
Location:
Round Rock, TX
senna,

I don't mean to rebuke, I mean to encourage.

Especially the way you have it layed out, it IS a very elegant spar...

Dan

18. Mar 29, 2003

### Qazulight

#### New Member

Joined:
Mar 29, 2003
Messages:
3
0
Location:
South east Texas
Senna,

I have an aluminum tube spar on my T-II. I would like to strenthen it without a big wieght penalaty. I was wandering if I could slip the Graphlite™ rod. into the tube then fill the space with that Great stuff foam you can get at the building supply.

19. Mar 30, 2003

### dannicoson

#### Member

Joined:
Feb 27, 2003
Messages:
15
0
Location:
Round Rock, TX
Qaz,

I am not familiar with the aircraft you are wanting to modify, however there are several issues you have to consider on a mod like this. 1) Foam will not work to transfer the loads you are considering between the existing aluminum and the graphlite. You would need a proper adhesive, probably epoxy based. 2) NEVER let carbon fiber connect with aluminum, even with a thin film of epoxy between. You would at least need a layer of glass between. Carbon and aluminum love to corrode. 3) the thermal expansion coefficient is quite different between these two materials. If you bond them the full length of a spar (which you would have to to here) you will build up a large shear between the two materials putting greater stress on the adhesive.

What is the spar configuration stock? Would it be easier to find a slightly beefier aluminum spar replacement? Why do you need to beef up this spar anyway?

Dan nicoson

20. Mar 30, 2003

### Qazulight

#### New Member

Joined:
Mar 29, 2003
Messages:
3
0
Location:
South east Texas
The 1983 model of the T-II uses an dual tube dual spar wing. The fore and aft spars make the leading and trailing edge of the wing. This is a typical ultralight construction. Each spar is made up of a 2 inch tbue sleeved over a 1 7/8" tube. the spars are 15 ft long. Aluminum tube can only be purchased in 12 foot lenths. By leaving the tubes not quite together you can get a 15 ft spar. the way the bending moments work the doubled portion of the tube carries the high bending moments and the single portions carry the lower stresses. It really is pretty clever.

The problem with the T-II is it is really fat. The cockpit is very wide and strong. The wings and motor are not. The plane was converted to a single place design because the gross wieght really can't be over 750 lb. The wing will take normal catagory g loading at this wieght, but there is no calculated safety factor and no room to gain weight. If the 50 hp rotax would get two people and fuel off of the ground, it would be over 750 lb gross and of course structurally unsound.

The only reason that there have been no known structural failures of this aircraft is it just will not lift that much wieght.

If I could increase the strenth of the wing I could install a rotax 582 and carry a passenger. As the plane is out of servcice now waiting for new fabric or sails, thought I might as well rebuild the wing.

I may go ahead and use a different airfoil as the one that is one the aircraft is horribly ineffecient.