TECHNICAL FIELD
[0001] The present invention relates to Cu-Zn-Al(6%) shape memory alloy having a low martensitic
transformation temperature and a process of lowering the martensitic transformation
temperature.
BACKGROUND ART
[0002] Cu-Zn-Al Shape Memory Effect (SME) alloys are promising smart and intelligent engineering
materials. (Wayman C.M., Journal of Metals, 32 (June 1980), p-129-137 and
[0003] Michael A.D & Hart W.B Metal Material Technology, 12(1980), p-434-440. These have
attracted much attention because of their low cost and ease of fabrication relative
to nitinol (White S.M., Cook J.M. & Stobbs W.M, Journal De Physique, C4 (ICOMAT-82),
P-779-783. But nitinol has superior properties, long fatigue life and is biocompatible.
[0004] There are about twenty elements in the central part of the periodic table Golestaneh
A.A., Physics Today, (April 1984), p-62-70 whose alloys exhibit shape memory like
Ag-Cd, Au-Cd, Cu-Al-Ni, Cu-Al-Mn, Cu-Au-Zn, Cu-Sn, Cu-Au-Sn, Cu-Zn, Cu-Zn-Al, Cu-Zn-Sn,
Cu-Zn-Ga, Cu-Zn-Si, In-Ti, Ni-Al, Ni-Ti, Fe-Pt, Fe-Pd, etc. (Wayman C.M., Journal
of Metals, 32 (June 1980), p-129-137 and Michael A.D & Hart W.B Metal Material Technol.,
12(1980), p-434-440.
[0005] Shape memory alloys (SMA) have a unique property i.e. these materials remember their
past shapes/ configurations. The Important characteristics of these alloys are their
ability to exist in two distinct shapes or configurations above or below a certain
critical transformation temperature. It undergoes diffusionless martensitic transformation
Golestaneh A.A., Physics Today, (April 1984), p-62-70, which is also thermo elastic
in nature i.e. below the critical temperature a martensitic structure forms and grows
as the temperature is lowered, whereas, on heating the martensite shrinks and ultimately
vanishes.
[0006] The martensite in shape memory alloys is soft in contrast to martensite of steels.
Deformation of these alloys is not by slip, twinning or grain boundary sliding but
by growth or shrinkage of self-accommodating, multi-oriented martensitic plates/variant
Saburi T., Wayman C.M, Takala K & Nenno S., Acta Metallurgica (Jan 1980) P-15.
[0007] On heating, the strained martensite reverts back to its parent phase, thereby, the
original undeformed shape is recovered. The change in structure can be linked with
change in shape and dimensions and the alloy exhibits a memory of high and low temperature
shapes. There is a usable force associated with these shape changes and thus the alloys
can be incorporated into range of temperature sensitive devices for warning, control,
detection, regulation etc .The actuators can be calibrated to operate within a narrow
temperature range by incorporating a compensating bias spring. The recoverable strain
is 2-8% and is dependent upon one or two way memory. Copper based shape memory alloys
in addition to one-way memory also exhibit two-way memory behavior, after undergoing
a suitable thermal-mechanical processing called training (Wayman C.M., Journal of
Metals, 32 (June 1980), p-129-137 and Michael A.D & Hart W.B Metal Material Technol.,
12(1980), p-434-440.
[0008] Once trained, the material will spontaneously change its shape when heated or cooled,
above or below the respective transformation temperatures. Forward and reverse martensitic
transformation temperatures are designated as 'Ms' (while cooling) and 'As' (while
heating) respectively.
[0009] In Cu-Zn-Al ternary alloys, shape memory effect lies in the Copper rich corner of
the triangle in the form of a trapezium. On enlarging this trapezium we can correlate
composition with martensitic transformation temperature Schetky L.M., Scientific American,
241 (Nov. 1979), p-68-76.
[0010] The amount of Aluminum varies from 4-10%, Zinc from 10-30% and balance is Copper.
Martensite formation temperature ('Ms') varies from -100°C to +300°C as a result of
very small change in composition. But, useful range for Aluminum brass is -70°C to
+150°C.
[0011] Martensitic transformation temperature (As) is extremely sensitive to composition.
A slight variation of either of the elements, Zinc or Aluminum (say ± 0.5%) shifts
the transformation temperature by ± 50°C. Therefore close control of composition is
utmost essential to get the desired transformation temperature for the actuator to
work at a specific temperature. Loss of low melting and volatile elements like Al,
Zn etc. while melting cannot be avoided in air melting furnaces. Vacuum melting furnaces,
in which close control of composition is possible but their installation is extremely
costly and are unaffordable to the small and medium scale melting units/industries.
[0012] In air melting furnaces, there is always a danger of loss of such elements inspite
of compensating these losses and following the necessary precautions rigidly during
melting.
[0013] The alloy with off-composition and undesired martensitic transformation temperature
has to be rejected or remelted. The efforts & inputs, thus put in, go waste. It was
also observed that loss of zinc or aluminum raises the martensitic transformation
temperature whereas increase of these elements lowers the transformation temperature.
[0014] Hence, the present invention is directed towards increasing or decreasing of martensitic
transformation temperature.
[0015] In Cu-Zn-4% Al alloy Adnyana D. N., Wire Journal International, (1984), P-52-61,
lowering of martensitic transformation temperature has been comparatively low i.e.
around 20°C-25°C.
[0016] There is always evaporation of volatile and low melting elements like zinc, aluminum,
tin, lead etc during the melting of copper base and other alloys, especially in the
air melting furnaces. These losses cannot be avoided but can be minimized by taking
all the care during melting, adding precisely weighed quantities of each element,
compensating for the elemental losses and rigidly following precautions during melting.
[0017] Vacuum furnaces precisely control these losses but their installations are costly
and are thus unaffordable to the small and medium scale melting/ foundry units. Cu-Zn-Al
shape memory alloys (SMA's) are no exceptions to these. The martensitic transformation
temperature (As) is an important parameter in shape memory alloys and is extremely
sensitive to the composition. A slight variation of either zinc or aluminum (± 0.5%),
as a result of melting losses, shifts the martensitic transformation temperature by
± 50°C. The material thus cast and processed reduces to a scrap & has to be remelted
thereby resulting in wastage of efforts, manpower and machinery.
[0018] Experimental studies show that it is possible to raise As temperature by 15°C-20°C
by the use of either a compensating bias spring or by selective etching /leaching
out of zinc by thermal treatments. But lowering of As temperature, once obtained poses
problems.
[0019] US Patent No.4, 634,477 recites about shape memory alloys. However, this patent does
not mention about the reduction in martensitic temperature.
OBJECTS OF THE PRESENT INVENTION
[0020] The object of the present invention is to provide for an improved process, in order
to lower the transformation temperature of Cu-Zn-6Al shape memory alloys, by a low
temperature re-betatising treatment from 110°C to 30°C i.e. a lowering of 80°C.
SUMMARY OF THE PRESENT INVENTION
[0021] The present invention relates to a process according to the appended claims.
DETAILED DESCRIPTION OF THE PRESENT INVENTION
[0022] The present invention provides for a process for lowering the Martensitic Transformation
Temperature(As) of shape memory alloy as claimed in claim 1.
BRIEF DESCRIPTION OF THE ACCOMPANYING PROCESS FLOW SHEETS, METALLOGRAPHS, BAR CHARTS
AND CURVES
[0023]
Fig 1 Shows experimental flow sheet of a process of production of Shape memory alloy in
the sheet form and its betatising (memorizing) heat treatment .It also depicts its
structure, SME response & martensitic transformation temperature. This process is
not according to the invention
Fig 2 Depicts microstructures of material betatised at 750°C/3 min/CWQ.
Fig 3 Depicts microstructures on heating the betatised material at various temperatures
like 200°C, 300°C, 400°C, 500°C, 600°C & 700°C.
Fig 4 Shows microstructures of seven more betatised samples, reheated (rebetatised) at
550°C, 575°C, 600°C, 625°C, 650°C, 675°C, 700°C (increments of 25°C) for ten minutes
and cold water (room temperature) quenched.
Fig 5 Flow diagram explains in details the condition of material, its microstructure and
shape memory response on heating the previously high temperature betatised material
at 200°C, 300°C, 400°C, 500°C, 600°C & 700°C.
Fig 6 Flow diagram explains in details the condition of material, its SME response, martensitic
transformation temperature (As) and its microstructure on low temperature re-betatising
of the previously high temperature betatised material at 550°C, 575°C, 600°C, 625°C,
650°C, 675°C & 700°C.
Fig 7 shows bar chart explaining re-betatising temperature Versus martensitic transformation
temperature (As).
Fig 8 shows the curve explaining re-betatising temperature versus martensitic transformation
temperature. It also depicts optimum lowering of martensitic transformation temperature
(As 80°C) on re-betatising at 575° C.
EXAMPLES
[0024] The following examples are provided to illustrate further the invention. However,
the same shall not be considered to limit the scope of the invention.
EXAMPLE 1
[0025] The charge consisting of commercially pure Copper, Zinc & Aluminum was melted in
an induction furnace under a charcoal cover and cast into sand moulds in plates of
sizes 150 x 100 x 12.5 mm. These were then homogenized at 800 °C for two hours and
cooled. These were then surface machined to remove oxidized layer. These homogenized
plates were analyzed for chemical composition. The plates (12mm thick) were reheated
at 750°C for one hour and hot rolled down to one-mm thick flat sheets with number
of reheating in-between the reduction passes. These sheets were held in fixtures (1.0-mm
thick sheets) and were betatised at 750°C for 3 minutes and then cold (ordinary) water
quenched. These were trimmed to desired dimensions, approximately, 20-25 pieces of
size 100mm X 10-12 mm X 1mm. From one flat sheet (betatised strip) a small rectangular
piece (10 x 10 mm) was cut and mounted in a acrylic compound, polished on grades of
silicon carbide papers then on diamond paste impregnated microcloth rotating wheel,
etched in potassium dichromate etchant and its micro structure was seen under the
optical microscope. The structure was fully martensitic. On the remaining strip, shape
memory response was seen by hot air blower. Transformation temperature was determined
using hot & cold water and temperature indicator. Its Shape Memory response was good
and transformation temperature was around 110°C-112°C (Fig.-1) Six more memorized
(betatised) sheets were then heated at 200°C, 300°C 400°C, 500°C, 600°C & 700°C for
ten minutes & cold (ordinary) water quenched. These were deformed to check the shape
memory response and their microstructures were analyzed (Fig.-2&3). The samples heated
to 200°C, 300°C, 400°C were very stiff as such memory could 60not be noticed. Sample
heated at 500°C was soft but without shape memory. Their microstructures were seen.
Samples heated at 600°C and 700°C were soft and showed shape memory at low temperatures
& high temperatures respectively. Their microstructures were also seen. Martensitic
structure prevailed between 500°C to 700°C. Thus seven more betatised samples were
further reheated (rebetatised) at 550°C, 575°C, 600°C, 625°C, 650°C, 675°C, 700°C
(increments of 25°C) for ten minutes and cold water (room temperature) quenched. Their
microstructures were observed (Fig.-4) These were deformed and their S.M. response
& transformation temperatures were determined (Fig.-5,6). A betatised sample, rebetatised
at 575°C for ten minutes and water quenched was martensitic with enough alpha at the
grain boundaries & within the grains, showed good shape memory response and its transformation
temperature (As) was around 30°C. An initial temperature of 110°C was thus lowered
to 30°C by this two-step treatment, a drop of temperature of 80°C by this process.
Any intermediate transformation temperature can be achieved by selecting appropriate
re-betatising temperature. The process of re-betatising was repeated number of times,
to ascertain the reproducibility and for the confirmation of results.
EXAMPLE 2
[0026] In the Cu-Zn binary phase diagram Higgins R.A, Engineering metallurgy Vol.1 (1971),
P-312-339, alpha solid phase (α) exists up to 39% zinc content. This α-phase has face
centered cubic structure (FCC). It is ductile, malleable and cold workable. Above
39% Zinc to 50% zinc content a beta phase (β) appears. It has body centered cubic
structure (BCC). It is a hard phase and can only be hot worked. Above 50% zinc content
a complex, brittle and undesirable gamma γ phase structure is formed. Cu-Zn-Al is
a ternary alloy system. It is basically a Cu-Zn alloy system with an addition of 3
rd element Aluminum. The zinc equivalent of Aluminum is six that is. 1% Aluminum has
an effect similar to 6% zinc (1Al = 6 Zn) West E.G. Copper & its alloys (1982), P-98-105.
We can thus calculate the equivalent of zinc for the shape memory alloy of composition
(74.4% Cu -19.5% Zn - 6.1% Al) by applying the following formula Greaves R.H. & Wrighton
H., Practical Microscopical Metallography (1971), P-159-177. The equivalent of Zinc
is calculated to be 43%.

[0027] If we examine the binary diagram of Cu-Zn System, the 43% Zinc content lies very
close to or almost on the phase boundary regions of α and β i.e. in the β rich, regions
of the diagram. Keeping in view their conditions relevant experiments were carried
out. On heating, the alloy to 750°C the structural transformation is from martensite
to beta, as such, we will designate this heating treatment as betatising and the transformation
temperature as As (while heating) & Ms (while cooling) respectively. The betatised
material (750°C /3 min./CWQ) was soft & fully martensitic with a good SM response.
Its temperature was 110°C to 112°C. The previously high temperature betatised samples
(6 Nos.) were rebetatised at low temperature to 200°C, 300°C, 400°C, 500°C, 600°C
and 700°C for ten minutes and cold water (room temperature) quenched. Their microstructures,
shape memory response and martensitic transformation temperatures were determined.
Rebetatised 200°C material was martensitic. But was very stiff, as quenching from
200°C and as such did not show shape memory. Rebetatised 300°C sample had α + β little
martensite. It was stiff and had no memory. The morphology of α was rod or plate type.
Rebetatised 400°C sample too was stiff with no SME. It contained α+β structure and
α-Phase was within the grains. The rebetaised 500°C material was soft but had no SME.
Its structure was α+β very little martensite. It had very thin α-phase rim at the
grain boundaries, which had tendency towards globular form. The samples heated at
600°C and 700°C were soft and deformable and showed memory of low & high temperatures
respectively. These materials were fully martensitic but 600°C rebetatised sample
had little α-precipitated at the grain boundaries and within the grains as compared
to 700°C sample. The 700°C sample was comparatively fine grained. These materials
did not crack even on cold water quenching. By these specific experiments it was ascertained
that shape memory effect in this material was between 550°C to 700°C. Thus, for further
experiment seven betatised strips were taken and were subjected to re-betatising treatments
at 550°C, 575°C, 600°C, 625°C, 650°C 675°C and 700°C (an increment of 25°C) for ten
minutes and then cold (room temperature) water quenched. Microstructures, shape memory
response, and transformation temperatures were evaluated. The 550°C betatised sample
was soft and its transformation temperature had dropped from 110°C to 22°C The sample
had a feeble memory mainly because of separation of sufficient volume fraction of
α-phase in β and very little visible martensite. Grain boundary α-envelop was also
thick. Sample rebetatised at 575°C was soft and had good shape memory 30°C. Its microstructure
was martensitic with enough volume fraction of α phase-streaks at the grain boundaries
and within the grain. α- Phase had tendency towards globular or lenticular shape formation.
The precipitation of α-phase from the matrix has enriched the remaining beta phase
in zinc content & shifted the composition towards right in binary diagram. This zinc
rich beta on quenching transforms to zinc rich martensite and thus lowers the martensitic
transformation temperature considerably i.e. from 110°C to 30°C a drop of 80°C. On
re-betatising at 600°C the material was soft with good SME and the transformation
temperature was around 45°C. The sample was fully martensitic with little α-phase
precipitated at grain boundaries & within the interior of grains. Samples rebetatised
at 625°C was also soft and has good SME around 61°C. Structure was martensitic with
unresolved α-phase at the grain boundaries. The sample rebetatised at 650°C, 675°C
and 700°C were all soft and had shape memory at 79°C, 100°C & 110°C respectively (Fig.-7,8).
These were fully martensitic and there was hardly any α visible at grain boundaries
& within the grains. In other wards α-phase was not resolvable. Since very little
or negligible α-phase has separated from the martensite matrix, these samples showed
shape memory at high temperatures. The results, therefore, indicate that the sample
rebetatised at 575°C gave an optimum value i.e. its martensitic transformation temperature
was around 30°C i.e. a drop of As from 110°C to 30°C which is a drop of 80°C by this
particular re-betatising treatment.
EXAMPLE 3
[0028] Cu-Zn-Al Shape Memory alloys (4% Al & 6% Al)
(a) 74.4% Cu-19.5% Zn-6.1% Al (Melt No. 7)
BETATISED (As): 110°C, REBETATISED (As): 30°C
(b) 74.1% Cu-19.5% Zn-6.4% Al (Melt No.5)
BETATISED (As): 130°C, REBETATISED (As); 50°C
(c) 73.6% Cu-20.2%Zn-6.2% Al (Melt No 6)
BETATISED (As): 83°C, REBETATISED (As)-10°C
(d) 71.0% Cu-24.8% Zn-4.2%Al (Melt No35)
BETATISED (As): 65°C, REBETATISED (As): 45 °C
EXAMPLE 4
[0029] It is observed that in Cu-Zn-Al (6%) shape memory alloys lowering of martensite transformation
temperature is substantial i.e. 70°C-80°C. It is also observed during experimentation
that martensitic transformation temperature could be raised by 15°C-20°C by incorporating
a bias or by suitable thermal treatment for the selective etching/loss of Zinc.
[0030] Martensite transformation temperature (As) can be lowered substantially by about
80°C in Cu-Zn-6% Al alloys by specific thermal treatment i.e. by low temperature re-betatising
of the previously high temperature betatised material. The decrease of temperature
was mainly due to the separation of small quantities of alpha (α) from the matrix
of martensite.
[0031] This retained α-phase does not affect the shape memory response but in turn, it assists
in cushioning the grain boundaries & thereby preventing the material from cracking,
even on quenching in cold water.
ADVANTAGES OF THE INVENTION
[0032]
- (1) In Cu-Zn-4% A1 alloy Adnyana D. N., Wire Journal International, (1984), P-52-61,
lowering of martensitic transformation temperature has been comparatively low i.e.
around 20°C-25°C whereas in Cu-Zn-6% A1 Shape memory alloys it was found to be substantial
i.e. 70°C-80°C.
- (2) The present improved process has no adverse effect on shape memory response, recovery,
fatigue life etc. rather the precipitation of α-phase in the martensite phase assists
in cushioning the matrix and generally associated with shape memory alloys.
- (3) The present improved process is likely to assist, the small & medium scale melting
units, to accept the risks and challenges faced, in the melting of shape memory alloys
by way of savings in the form of cost, manpower and machinery.
- (4) The process is novel, simple & needs no additional manpower or equipment.
- (5) Cold water quenching gives better shape memory response.