Composite tubing material properties 1

[bsdhr]

Hello Everyone;

I am putting together weight saving proposition for upcoming Boeing 787 part. The intended design is all Stainless Steel and the proposed design would be combination of Steel and Composite material.

We are currently the selected supplier for this part and hence have all required and generated engineering information to design the part. However, we don't have work expereience with composite materials. So help/suggestions on below would be greatly appreciated;

1.Part made out of composite material would be in form of round tube about 12 inches long. The outside diameter is 1.10in and inside diameter is obtained from stress calculations. I need material properties, to simulate via hand calculations axial, bending and torsional stresses.

2. Preliminary work scope for structural stress calculations would assume linear elastic isotropic properties. This is for simplicity as I understand composite layup can be tailored.

3. My calculations on current design indicates, 50% of Stainless Steel material can be removed without affecting functionality. However, lost material has to be compensated by composite tube to account for lost structral strength. If the weight savings upto 25% can be justified without loss of structural strength, i will like to procure few pieces to build prototype. So please suggest something off-the shelf for low cost concept feasiblity check

Again thanks for taking interest in my quest.

Bye.

 

[SWComposites]

What is the function of the part?

What is the cert basis: primary, secondary, tertiary structure?

What are the loads on the part?

Assuming isotropic properties may not be appropriate, depending on the loads.

Material properties can be found in Mil-Handbook-17, Volume 2. Buy a copy at

http://www.astm.org

What do you mean by combination of steel and composite? how joined?

What manufacturing process do you plan to use for the composite part?

Composite tube suppliers can be found by a Google search

 

[bsdhr]

Thanks SWComposites.I will try to answere as much as possible.

What is the function of the part?
The part acts a cantilever beam. Sorry, I cannot tell you exact part name or application. So I will restrain to physics of the part in description.

What is the cert basis: primary, secondary, tertiary structure?
Current cert is secondary. Although similar part, obvisouly different design, is certified as primary.

What are the loads on the part?
Being cantilever, the part sees bending moment, axial thrust and torsional load. Axial Limit load for instance is 25,000 Lbf.

Assuming isotropic properties may not be appropriate, depending on the loads.
Yes, you are right. But my approach is to start with simpler models and migrate to complicated material properties in future iterations.

Material properties can be found in Mil-Handbook-17, Volume 2. Buy a copy at

http://www.astm.org

Thanks for the suggestion.

What do you mean by combination of steel and composite? how joined?
The actual composite laying process would involve laying of fibre layers and injection of epoxy. But for begning, inside diameter would be machined to match the tube outside diameter and epoxy to fill any radial gap.

What manufacturing process do you plan to use for the composite part?
Manufacturing process would be defined by supplier and approved by Boeing.

Composite tube suppliers can be found by a Google search.
I am googling too.

 

[GBor]

I would say you are headed down the wrong track. Use a pultruded or filament wound carbon-epoxy with fibers laid along the axis and at +/-45 to handle the torsion. I would recommend Engineering Mechanics of Composite Materials by Isaac Daniel and Ori Ishai as a good starting point.

The bigger problem here is failure mode. It isn't like homogeneous materials. You have matrix cracking, fiber failure, interface issues, etc. etc. The stiffness matrix can be as much as 36x36 for multi-layer, multi-directional composites compared to Young's Modulus for isotropics.

Have fun on the quest, but be careful. The guys at

http://www.materials-sciences.com

have some online calculators that may help you as well (including a material property estimator).

Garland E. Borowski, PE
Engineering Manager
Star Aviation

 

[SWComposites]

Well, you are starting out wrong. With composites, it is vital to select the manufacturing process at the same time as the material is selected and the design defined. The design/material/process are all linked, and to treat them separately is likely to result in a failed project (been thru this way too many times).

Do not machine the thickness of the composite (I assume this is what you mean by machining to match the diameter).

You will likely need an adhesive to bond to the steel. Standard epoxy matrix resin is likely not adequate. Need to select the adhesive material early also. And surface prep of the steel will be critical and may drive the manufacturing process for the bond. The FAA (and Boeing) will get real excited about any bonding and will ask you lots of questions.

 

[RPstress]

While it's right to bear in mind the additional complexity of composites over metals, you can still do preliminary sizing without knowing the manufacturing details. You can size the part with some tentative allowables.

For secondary structure it's probably appropriate to design to 'open hole' allowables. This might translate, for a modern woven HS carbon (e.g., AS4/T700/HTS) epoxy (e.g., 977-2/8552/M21) in a quasi-isotropic layup, to about 175 MPa (25 ksi) compression, 225 MPa (33 ksi) tension and 120 MPa (17.5 ksi) shear, allowing for hot/wet conditions.

If you can get away with a 0/45/0 type woven layup (depends on torque vs. bending/axial) the allowables would be more like 185 MPa (27 ksi) compression, 275 MPa (40 ksi) tension and 100 MPa (14.5 ksi) shear.

For primary structure the compression and shear allowables would drop by about 20%.

Depending on the bending you may well want to grade the composite thickness. This takes a bit of care, but is doable with common sense.

Stainless (even if it's reasonable stuff like 15-5PH or good stuff like PH13-8Mo) will work out heavier than titanium alloy. There is also more experience on bonding to Ti than stainless.

MIL-HDBK-17 is definitely worth having, though it's huge hunk of data. A simpler ref. might be in order as well.

Official website:

http://www.cmh17.org/

To get an official copy for free:

http://assist.daps.dla.mil/quicksearch/

(this is the link cited on

http://www.cmh17.org/resources.aspx

)
Enter MIL-HDBK-17 in the 'Document ID' box.
On the next page click on the volume of interest.
On the next page scroll down to the bottom and click on the pdf symbol next to the revision of interest.

 

[rb1957]

it's secondary structure, carrying 25,000 lbs ??, with a maximum moment of 25,000 ft.lbs ??

don't quite understand para 3 in the OP ... is there a MS = 0.50 on the part ? what heat treat is the SS (160ksi, 200ksi, 75ksi) ? what is the thermal environment of the part ? shoock loads ? fatigue loads ? is there a space constraint ? (in my experience pretty much the only reasons for going to SS is because of a hot environment or severely limited space).

if the load is uni-directional, why not an I-beam ?

 

[RPstress]

Forgot to mention, if you have complex loading (axial + bend + torsion at the same time), the axial and bending will largely be carried in the 0 degree plies and the torsion in the 45s, but you can't quite have both present and still use the allowables I gave. The axial stresses will still cause some load in the 45s and the torsion will load up the 0s a bit.

For inital sizing if you've got significant torsion with axial/bending you should knock the allowables down by a good 10% and do a ply-by-ply analysis quite early (e.g., use The Laminator or other freeware/shareware with a Tsai-Wu failure criterion or similar). If you size with the allowables given you'll be being a bit optimistic (=dangerous).

 

[bsdhr]

G'morning folks!

It was pleasing to see my in-box crowded by your valuable replies. Thanks for your comments and I will attempt to respond individually to you as below;

GBor: Thanks for the link. I am refering to "Principles of Composite Material Mechanics" by Ronald F. Gibson. Took his masters level course 8 yrs ago at school. What do you think abou this book.

SWComposites: By no means I implied machining the composite part. What I was refering to was that material removed by machining the inside diameter of the exsisting SS part would be filled by the composite material to regain the lost structural strength. Yes bonding is definitly an issue.

RPStress: You sound like me! In todays world it's impossible to convince managment for investment into full blown engineering efforts. Everybody is trying to run their place at shoe-string budget. Plus I have silent resistance from engineering community in-house, besides being skeptic, they see vunerability by composites. It's difficult to change people, especially when they haven't seen change to manufacturing process of this part for ages. Please, by no mean I meant dis-respect to them rather merely expressing momentary frustation. Hence, I intend to pursue this challange in smaller baby steps.

rb1957:Yes the part has limit load requirement of 25,000lbf, but torque loading is not 25,000 ft-lbs. Their is torque load. 50% is the amount of SS material that can be removed from ID of the tube, but than has to be replaced by composite material to bring back the lost strength. What is going to be final ID dimension, ... I don't know till I calculate. And hence am looking for representative isotropic properties for composites. I didn't mean MS =.5. THe SS part is heat treated to 52 HRC. The part is encased in housing and not exposed to atmosphere, however is subjected to -65 deg C to +125 deg C thermal loading.

All, again what is being discussed here is merely a concept of using composite in combination with SS to provide a composite structure that meets or exceeds the design loads. I uploaded simple cross sectional sketch to help.

Thanks to all of you for your interest and replies.

Bye.

 

[rb1957]

i assumed a radial load, 25,000 lbs * 12", i assumed the cantilever you mentioned in your 1st post .... whatever.

what does "50% is the amount of SS material that can be removed from ID of the tube, but that has to be replaced by composite material to bring back the lost strength" mean ? how do you know "50% is the amount of SS material that can be removed from ID of the tube" ?

52 HRC = 283ksi steel ... that's pretty darn strong stuff.

something to note, is that the cantilever doesn't need to be a constant thickness, that the moment is increasing along its length.

something to remember about composites is that they're expensive to make, maybe saving a lb or two isn't worth the cost ?

 

[SWComposites]

Another good reference is the ASM Handbook, Volume 21 Composites. Can be purchased from

http://www.asminternational.org

http://asmcommunity.asminternational.org/portal/site/

http://www/menuitem.2b9d1953d012ee1...253501f0f8110VgnVCM100000701e010aRCRD&itemId=

 

[RPstress]

What is the stainless material?

Stainless as strong as 300M...I could use that!

 

[bsdhr]

RPstress: The material is pre AMS 5898.

 

[RPstress]

Thanks. That's quite nice stuff. Not one I'm familiar with.

 

[Hansmeister]

BSDHR,

I've done this problem many times. Suggestions:

1. Are there any deflection or stiffness requirements?
2. If not, stick with the T700 fiber and a current generation 350F cure resin system from your favorite prepregger (Cycom or Hexcel).
3. If OD is fixed, then solve for ID. As a Q&D, I came up with a wall thickness of .16 inches.
4. The part will be made on a male mandrel with vacuum compaction every 4 to 6 plies.
5. Autoclave cure if possible.
6. If the OD is critical, put a layer of 5HS on last and grind to size.
7. Use Ti fittings and a good Hysol adhesive.

 

[bsdhr]

Hansmeister;

I am glad you came across my thread and shared your valuable expereience. I did hand calculations and based on fixed OD of 1.125 in, for an ID of 0.125in (say) I get 25 Ksi of normal (axially directed) stress and 7 Ksi shear stress when a combined limit load ( axially compressive) of 25,000 Lbf and torque load of 2000 lbf-in is applied.

Based on generalized composite material properties given by RPStress, the design would fail. But your rough calculations contradict this. Can you please share more light on this with me?

Thanks.

 

[rb1957]

OD = 1.125"
ID = 0.70"
A = 0.61in2
I = 1.07in4
rho = 1.325in
L/rho = 16 (very short column)
P = 25000 lbs (axial compression)
stress = 41ksi
T = 2000 in.lbs
shear = 2000/(2*pi/4*1^2*0.125) = 7ksi

why not Al (weight < 1lb) ?

 

[Hansmeister]

I have run a quick sample problem for you to consider.

Assume T300/Hexcel 74 resin, .005 cpt, and a repetitive layup sequence of [0-2/±60/0-2]n, 6 plies per set at .030" thick, repeat as necessary to achieve desired load balance.

Ex=14.16 msi, Ey=4.88 msi, G-xy=1.82 msi, and v = .308

The strength allowables (failure) Tsai-Wu criteria:

Ftu-x=133.3 ksi, Ftu-y=39.3 ksi
Fcu-x=-129.7 ksi, Fcu-y=-33.0 ksi
Fsu-xy=±19.4 ksi

Apply appropriate safety factors on design and fabrication process.

 

[rb1957]

guys, what's the cost difference between composite and Al ?

there seems to be little reason that an Al part wouldn't weigh more than 1 (maybe 2 lbs) ?

no idea why it was 280ksi SS in the 1st place ??

 

[bsdhr]

Good morning folks!

With all the help I am getting on this thread, I personally feel bad for not sharing with you the application and part usage on plane. But you have to forgive me on this, the simplicity of the application would allow anyone to comprehend the idea and leap-frog my employer in first to market. I hope you understand my interests.

RB1957: With above disclaimer, I have to evade answering your latest question ( i.e. "why 280 ksi SS"). Al is very good idea and I don't discount it.

Hansmeister: Thanks for sharing data.