Determine Ti-6Al-4V orthotropic properties with NDT?

[dculp1]

I have a specimen of Ti-6Al-4V plate (approx. 3" x 3" x 5" with the grain in the 5" direction). I need to determine the 9 orthotropic properties (3 Young's moduli, 3 shear moduli, 3 Poissons ratios). However, I don't want to alter the specimen (e.g., for tensile tests). Can the orthotropic properties be determined by measuring the wave speed in the three directions with ultrasonic pulse echo? If so:
- Who could perform these tests?
- Where can I find literature that gives details and limitations of this test?

Are there other non-destructive tests that would give the properties?

Thanks,
Don Culp

 

[metengr]

I know of no such credible test methods. I would rather search for and use published mechanical property values. This material has been investigated and is commonly used in industry and aerospace.

 

[dculp1]

I have contacted metallurgists at several Ti manufacturers (TIMET, RMI Titanium, etc.) Their response has been that they don't have the orthotropic properties. They suggested that aerospace companies may have these values (but, of course, such information would be confidential). A reasonably extensive on-line search has not yielded any useful results. (Isotropic properties are generally given (e.g., per matweb.com) but no orthotropic properties.)

Also, the properties can vary substantially depending on the manufacturing process (e.g., forging versus extruding), the material's form (sheet, rod/bar, plate), and the heat treatment. Hence, I need a method to determine the properties of my particular sample. Since I may need to run additional tests on my specimen I can't sacrifice this specimen to make tensile test specimens.

Thanks for any additional suggestions.

 

[mrfailure]

This might be a good question to also post in the Industrial/Manufacturing section in the QC, Inspection & Testing engineering forum.

Aaron Tanzer

http://www.lehightesting.com

 

[metengr]

Might be related to your topic;

http://www.scielo.br/scielo.php?pid=S1516-14392011000300008&script=sci_arttext

 

[metengr]

and this paper

http://www.gruppofrattura.it/ocs/index.php/gigf/YSESM2010/paper/viewFile/888/799

 

[dculp1]

I am somewhat familiar with using electronic speckle pattern interferometer (ESPI) with FEM to determine orthotropic material properties. For instance, see the article by Lecompte et al (Identification of Elastic Orthotropic Material Parameters Based on ESPI Measurements) --

http://orbi.ulg.ac.be/bitstream/226...pic parameters based on ESPI measurements.pdf

In this article (and in Barile's article that you reference), thin plates are used. For this case of plane stress only 4 material properties need to be adjusted in the FEM in order to get agreement with the ESPI (see page 6 of the above article). For the general 3D orthotropic case, 9 material properties must be adjusted which makes the process very much more difficult. Do you know anyone who uses this method for the general 3D case? I would be particularly interested in how the specimen is loaded for the ESPI and the sequence by which the properties are adjusted in the FEM.

 

[Maui]

Unlikely. But the following paper may give you a direction in which to go, since it does involve the alloy and geometry that you stated:

http://www.springerlink.com/content/73817m53671m2742/

You may find Fig. 2 in the following link relevant as well:

http://www.worldscibooks.com/etextbook/4311/4311_chap01.pdf

Maui

http://www.EngineeringMetallurgy.com

 

[dculp1]

Maui --

In your first link "Calculation of the Elastic Anisotropy of Ti:6Al-4V..." there is an interesting statement in the full text preview (just above the table): "The elastic constants for single crystal titanium have been determined ultrasonically by Flowers." Hence, an ultrasonic method may be possible.

 

[Maui]

dculp1, it is possible for single crystals. The elastic constants for this material will vary depending upon the crystallographic direction in which they are measured. If you have a single crystal and know the orientation, then the pulse-echo technique should provide a relatively simple and accurate method for determining the elastic moduli.

But if the material you have was produced using standard industrial melting and processing techniques, then it is composed of a large number of radomly oriented grains. Using the pulse echo technique on such a material will provide you with an average value of the elastic constants over these various crystallographic orientations. For this reason, you won't be able to determine their values in specific crystallographic directions for the material that you have in hand using this technique.

Maui

http://www.EngineeringMetallurgy.com

 

[dculp1]

Maui --

I don't need the properties in specific crystallographic orientations. I need the properties in the longitudinal, long transverse, and short transverse material directions (i.e., the same as would be obtained using standard tensile tests on an orthotropic material). I will use these properties in FEA/FEM.

With this consideration, could pulse-echo be used on my unmodified specimen? If so, do you know who could provide this service?

 

[Maui]

For an isotropic material, the elastic modulus is independent of direction and can be determined using ultrasonic techniques by simply measuring the wavespeed of a longitudinal pulse according to the equation

V = [E/rho]^0.5

where V is the wave speed, E is Young's modulus, and rho is the density of the material. Where the material is anisotropic, the relationships become more complicated, as described here:

http://www.ndt-ed.org/EducationResources/CommunityCollege/Ultrasonics/Physics/elasticsolids.htm

I'd suggest contacting a certified level III ultrasonic operator and posing this question to them. They should be able to do it, or suggest someone who can.

Maui

http://www.EngineeringMetallurgy.com

 

[metengr]

If you are dealing or assuming an isotropic material why not just use the published value of E, G and v. You don't need to go through all of this effort.

 

[Metalhead97]

dculp1,

The only way your going to get orthotropic properties in bar or plate stock is if you have a texture. It might be cheaper just to do an x-ray diffraction analysis by an outside lab on the material you have available to figure out the texture. Otherwise, I would try to measure the speed (as Maui said) of longitudinal waves in 3 directions.

Metalhead

 

[dculp1]

metengr -- Vibration measurements of this block in its fundamental longitudinal mode indicate that the material is significantly orthotropic (i.e., symmetric locations don't have equal amplitudes; I have confirmed that these differences are not due to measurement errors). Also, if I model this block with FEA using isotropic material properties, the FEA results don't agree with the measured amplitudes, regardless of how the FEA isotropic properties are adjusted. Hence, orthotropic properties are required.

metalhead -- I can measure the longitudinal wave speed in the three directions. However, how do I convert these wave speeds into the 9 material properties for an orthotropic material?

 

[Metalhead97]

dculp1,

Well the shear modulus should be determined analogously to Young's Modulus using shear waves. The only way to extract any info related to the Poisson ratio by ultrasonic tests is to send longitudinal waves each of the three axis while containing (dampen) the other two and measuring the speed. Of course, there are details that you will need to work out.

Metalhead