08:45 May-13-1997 Steffen Bonss Engineering IWS Dresden, Germany, Joined Nov 1998 1
Re: acoustic speed of liquid metal lock length under the bath. The interface between melted/solid titanium : should give the reflection, which could be used to measure the unmelted titanium length. : If the measurement is performed from the very begining, when all the titanium is not melted, : latter measurement can be subtracted from initial one. : Liquid titanium could change its volume, so the optical measurement can be done from the laser : beam side. In this case laser beam might be employed to do that. : One also has to be aware of raising temperature effect on solid titanium ultrasound speed. : If there is no liquid in the ultrasound path from transducer to melted/solid titanium interface, : dual wave transducer can be used. Interesting choice are Zinc oxide sputtered transducers. : If ZnO crystals are sputtered at some agle, such transducer is producing longitudinal and share : waves at the same time. In such case the temperature effect can be eliminated. It will also : produce a good acoustical coupling. But... How much of temperature these transducers culd withstand? : In case the temperature at measurement point is quite low, it couls be the case. : I believe there should be one more case considered. The bath will be of a hemisphere shape. : In case wide beam transducer is ised, there will be a distortion in bath bottom depth measurement. : I think the transducer should be foccused, with focal point located at liquid-gas interface. : In case there will be penetration of melted bath deeper to titanium block, the position of surface : might change, so the focal spot length should be under consideration too. : All said concerns the pulse propagation in the liquid titanium bath measurements. The transit-time : resonanse, tone-burst phase match and etc techniques can be considered.
: There is one more idea. The ultrasonic beam, after passing the liquid titanium bath, can create : the waves on the melted titanium surface. By laser interferometry measurements (might be shearography : can be suitable too), the spatial distance between two peaks can be measured. The wavelength can be : calculated in such way...
: Linas Svilainis
The cooled coupling plate is planned to make of an AlMg-alloy. The maximum temperature at the point of transducer contact shall be about 373 K. At the back site of the titanium plate the temperature raises to about 873 K, but very localized beneath the melting bath. The liquid titatium is in fact a pool on top of the solid titanium, but a pool wandering along the surface. While the transducer position is fixed beneath the titanium, the liquid pool moves through the sound field. The method suggested by Linas for measuring the melting depth should be possible, if the echos are measurable. The second way for measuring the melting depth is cutting and polishing the specimen after processing. The position of transducer and specimen, the track speed and the moment when the measuring starts are exactly known. In this way it should be possible to determine the melting depth too.
Re: acoustic speed of liquid metal : The cooled coupling plate is planned to make of an AlMg-alloy. : The maximum temperature at the point of transducer contact shall : be about 373 K. At the back site of the titanium plate the : temperature raises to about 873 K, but very localized beneath : the melting bath. The liquid titatium is in fact a pool on top : of the solid titanium, but a pool wandering along the surface. : While the transducer position is fixed beneath the titanium, the : liquid pool moves through the sound field. The method suggested : by Linas for measuring the melting depth should be possible, if the : echos are measurable. The second way for measuring the melting depth : is cutting and polishing the specimen after processing. The position of : transducer and specimen, the track speed and the moment when the : measuring starts are exactly known. In this way it should be possible : to determine the melting depth too.
: Steffen Bonss
Thanks for the update. My additional thoughts would be as follows:
The surface temperature of 373K/100C is beyond the safe limit for most standard, commercial broadband contact transducers. So the first consideration is that you will need a transducer that is capable of withstanding this temperature without failure. You should discuss this in detail with your vendor, as a special fabrication may be required to prevent thermal damage.
As an aside, I would probably not recommend a high temperature delay line transducer because it would introduce additional delay line echoes into the mix, and I would not use a high temperature dual element transducer because it would probably lack sufficient resolution for the precision echo-to-echo measurement required here.
We haven't discussed the typical depth of the liquid pool. Assuming it is a couple millimeters, I would try a 5 MHz broadband longitudinal wave transducer. Coming from the bottom, if everything is properly coupled and aligned, you would see an echo from the cooling plate/titanium boundary, probably a second multiple of that echo (based on the rough dimensions from the drawing, an echo from the solid titanium/liquid boundary, and then an echo from the top surface of the liquid. You will want to measure the interval between the last two echoes. Depending on the exact material thicknesses and gain levels required, you may also have a third multiple of the cooling plate/titanium boundary echo in here somewhere.
I suspect that experimentation with an actual assembly will be required to optimize both transducer selection and setup parameters. As a start, I would try coupling through the cooling plate/titanium assembly when it is cold to verify that I was getting a strong signal from the top of the titanium. Based on a total thickness of about 20 mm, and assuming good coupling, I think a 5 MHz broadband transducer would be a good place to start.
09:18 May-13-1997 Linas Svilainis R & D, Kaunas University of Technology, Lithuania, Joined Nov 1998 67
Re: acoustic speed of liquid metal : : : : : : : I would like to measure the acoustic speed in liquid titanium : : : : during laser gas alloying. The enclosed figure shows the intended : : : : arrangement. Can you give useful hints about suitable transducers : : : : and frequencies?
: The cooled coupling plate is planned to make of an AlMg-alloy. : The maximum temperature at the point of transducer contact shall : be about 373 K...
373K=100C? It should not be a problem-we could offer longitudinal wave foccused transducers in a 5-10MHz range, which have wide bandwidth (90%BW@-6dB). Of course, the transducer design should be adjusted bearing in mind the temperature and material of investigation.
: ... measuring the melting depth should be possible, if the : echos are measurable....
I am not familiar with titanium properties(attenuation, aproximate speed of melted media, how nasty this material is). Therefore I am not suggesting going higher frequencies. The 10MHz transducer will give you about 0.6mm wavelength. In such case the echoes from top and bottom of the bath will "stick" to each other(I assume your bath depth is of such order).Using first reflection "stripping" procedure and resonant investigation techniques this could be handled. (http://www.ultrasonic.de/article/wsho0597/linas/linas.htm#4) Nice results for multiple reflections case are in R.Freemantle paper (http://www.ultrasonic.de/article/wsho0597/freem/freem.htm) Paper of W. Hillger describe the technical aspects os similar problem: (http://www.ultrasonic.de/article/wsho0597/hillger/hillger.htm) The methods mentioned deal with thin composite plates, but the results obtained should be aplicable to your case too. Might be I am going too far and simple deconvolution procedures (for instance L1 norm or ASE) could handle this easily.
Basing the experience with other materials, ultrasound speed should go down substantially,which should ensure relevant reflection from solid/liquid interface. The other question,which I raise again is the influence of temperature of solid titanium. The TOF change due to temperature should be about 0.05% for 1C(I use steel data). If temperature is raised from 20 to 500C, then one has 25% 'float' in measurements. This should be compensated somehow.
Re: acoustic speed of liquid metal : I would like to measure the acoustic speed in liquid titanium : during laser gas alloying. The enclosed figure shows the intended : arrangement. Can you give useful hints about suitable transducers : and frequencies?
I suggest that you discuss your problem with C.K. Jen of the Industrial Materials Institute of the National Research Council of Canada. He is already doing ultrasonic measurements in other molten metals.
André Moreau Industrial Materials Institute 75 de Mortagne Boucherville, Quebec J4B 6Y4