Vous êtes sur la page 1sur 14

c 


The same two paint systems (Hempel¶s 15500 and 45880) were tested for tensile strength.
The elongation test was conducted for two different widths: 19mm and 27mm. Each of the
width were tested at three different thicknesses: 100ɰm, 200ɰm and 300ɰm., and each of the
thicknesses at three different speeds: 2mm/min, 10mm/min and 50mm/min. Each test was
carried out using five replicates or more. A schematic description of the experiment planon
the figure given.

The aimed gauge length of the samples was 100mm in the beginning, but it was changed to
just 40mm due to a problem with the paper used as surface material during painting and
drying. The paper had small dents/lines dividing it into four sections. These dents deformed
the samples, easing thereby breakage in the line of deformation. We were not aware of this
problem until we were finished making our samples. Almost all our samples had that small
deformation in the middle of the length. All deformed samples were therefore manually
broken at the deformation line. This breaking of samples added more replicates to the later
testing.
A few samples that was not deformed was used in their initial length. These were used in the
Sticks-Full size test.
Preparing elongation samples

Due to the tight schedule, we had to make 180 samples in only one day. That is why we had
to find a painting method that was efficient and fast. We considered two, in principle,
different methods- the ³cutting method´ and our so-called ³tape method´. In both methods
the paint was applied on paper that already had the right dimensions. In that way, we would
avoid having to cut in films that had been made already. Some earlier experiments in doing
the cutting after filmmaking has been conducted at Hempel with both good and not so good
results depending on the paint system (how brittle it is).

ß 
  : The method was used on most of the preparations. It was carried out by
cutting sticker-paper into the desired shape of the samples and placing them above a
platform, which in our case was a wooden stick. The paper was attached to the stick using
tape at both ends.

The wooden stick had two purposes. It had to elevate the sample while spray-painting and it
had to prevent the paper from bending when painted.

ã 

  
    


This method was very efficient because we were three people working in what looked to be a
production line. It will be very slow though if one works alone.

› 
  : Another way to prepare the samples was by dividing a large plate using tape
into sections. Removal of the tape immediately after painting causes the surface to split into
areas that has the same form as the desired samples. This depends of course of the way the
tape was placed before the spray painting. It is very crucial that the tape is placed correctly if
it is desired to have the same width throughout the whole sample. The method is even more
time demanding than the cutting method.
ã 

 
    

When making the samples by the tape method, one must make sure that the paper is held in
place as explained before.
To ensure that the sample sticks made by the two methods can be differentiated we use the
term sticks for the ones made by the cutting method and tapes for the ones made by the tape
method.

After spray painting, the tape strips was immediately removed, and the paint was left to dry
for 24 hours at room temperature (20,4°C) (both the sticks and the tapes).

ã 

  
      
 

Then removed from the paper and cured/dried 48 hours in the oven at 40°C and 24 hours at
60°C. The reason for the temperature change was to boost the curing process. The samples
should have been in oven at 60°C over weekend but by mistake the oven was not switched on
correctly.
‰easuring
The samples were taken out from the oven and placed in a climate room for 12 hours before
measuring. The climate room was kept at a temperature of 23,5°C and with a humidity of
50%.
To measure the elongation, a Zwick tensile testing machine was used.

Two clamps were used to hold the sample during elongation; one at the top and one at the
bottom. The clamps were squeezed together using a force equal to 2 bar.

Placing the sample correctly was difficult if the sample was narrow. It had to be placed
vertical between the clamps so that the elongation will take place in the long direction only. It
was a subjective judgment whether or not the sample was correctly placed. This could have
caused an error in the report. The maximum force would be influenced by this error.

The force used to pull the sample apart was measured. Both the maximum force and the force
at break were printed out in a report (Fmax = Fbreak in 99,5% of the cases). By dividing the
maximum force by area before elongation of each sample¶s cross-section one obtains the
engineering tensile strength. And by dividing the force at break by area before elongation of
each sample¶s cross-section one obtains the engineering stress at break

Also the strain in mm was measured and printed out in the report. The strain in was given in
percentage as well. A graph showing the force as function of the strain in percentage was also
included in this report.

The sample length was used by the computer to calculate the strain. Unfortunately one of the
Zwick software weaknesses was pronounced in all our reports as it did not calculate the strain
based on the manually entered gauge length. Instead the length was based on a default length
that had to be entered in the program¶s ³wizard´. A normal user would think that it was
logical that the calculation should be based on data entered on the screen, but Zwick¶s
software demands from the user to run a wizard. This could have been acceptable if it was
stated clearly in the user manual, which was not the case. Apparently, nobody was aware of
this and wrong calculating were been accepted all along.

The strain calculation in all our reports was based on a length of 150mm. We calculated the
real values of the strain and strain percentage manually in Excel, and we added a real ³scale
to all our reports´. The new scale showing %-strain, is placed underneath the original one.

?
?

?
?
?

Elongation test
When treating the data from the tensile test, hundreds of plots could be made. Because of big
variance of the tensile strength, Fmax, and the %-strain the focus will be on the modulus
obtained at 0,1mm elongation, which turned out to be very consistent.
General observation
After plotting the thickness as function of modulus, one very consistent information is
obtained. Even though this was not expected, the thickness has an influence on the modulus.
This influence is observed for both paints and on all samples .It seems to have the same
degree of impact on the two paints making the slopes about the same for the two paints.
Where a modulus in the area of 3000 ‰Pa is found for a thickness of 300m m the modulus
increases significant to about 5000 ‰Pa when the thickness changes to 100 m m (15500
paint).
This phenomenon will be discussed below where several possible explanations are listed.

? The apparatus used for elongation test


? 6ength of samples
? Curing time of paint

þ 
When performing a mechanical measurement like the tensile test, it is always related to an
uncertainty of some degree. Because of many moving parts inside the machine, some errors
will occur. The measurement of the force used to elongate the sample is measured using a so-
called strain gage. A strain gage is designed to convert mechanical motion into an electronic
signal. A change in capacitance, inductance, or resistance is proportional to the strain
experienced by the sensor. Even though the elongation of the strain gage is very small, it will
lead to an error in the measurement. When the sample, as in this situation, is very short, the
error coming from the strain gage becomes larger.
Unfortunately, it has not been possible to find any numbers about these uncertainties in the
instruction manual for the tensile testing apparatus.

.
6  

As mentioned earlier, the length of the samples turned out to be smaller than wanted because
of problems with the paper used for the filmmaking. Due to these small samples, the error
within the apparatus will be pronounced. When placing the film between the clamps in the
tensile tester, it is now of even bigger importance that the film is in total vertical position,
since unwanted forces from the edge would break the film more easily. However, this type of
error has not been seen. Actually, the opposite is observed. The short films had surprisingly a
longer %-strain than the long samples and hence do not seem more brittle as expected if
placed wrong. The short films also breaks at higher forces, Fmax, than the long ones. An
explanation to this can be, that the longer the sample, the bigger the chance to have defects
within the sample; especially along the edges as in the figure below.

ã 
          
 
When stretching a sample, some degree of biaxial stress will be transferred to the sample.
The smaller the ratio between length and width, the larger the biaxial stress. However, the
results do not indicate this phenomenon since the short samples elongates more than the
longer ones. Unfortunately there is no minimum or standard length given for the tensile
tester. After looking at dimensions of some metals for tensile testing, a recommended ratio of
6 between length and width was given. The ratio used for these experiments were:
(40 x 27) mm = 1,5; (40 x 19) mm = 2; (120 x 27) mm = 4,5.

ß 
 
If the paint has not cured enough, different phases will be present in the film. If the surface
layer is stronger than the inner layer, then plots of thickness vs. modulus will indeed have a
negative slope as found in the results. The smaller the thickness, the more dominating the
surface layer will be.

The hardness test indicated a difference in the curing of the two paints. As expected the
15500 was the most brittle one. The hardness test conducted on the 45880 showed a
correlation with the film thickness. The thicker the film, the more soft the material. This was
observed on all the samples ± also on the plates which had been exposed to higher
temperatures for several days. If the curing time for the 45880 was not long enough, this
could explain the lower modulus at higher thickness. However, the correl ation between
thickness and modulus is also seen on the 15500 system without having varying hardness
when thickness changes.
Change of parameters during elongation
In the following discussion, the focus will be on the various changes of parameters.

ß
    

By looking at the different plots of the thickness as function of modulus, a difference through
all the plots is observed. As expected, the one 15500 has a higher modulus than the 45880.

ß
  
Before measuring, you could perhaps expect that the modulus from the paint with dimension
40 x 27 mm would be more brittle than the same paint with dimensions 40 x 19 mm because
of larger biaxial tensions within the film caused by the low ratio between length and width.
This was not observed.
Instead, the width seems to have a slight effect on modulus for the 15500 paint system only -
at all speeds. A theoretically explanation for this behaviour cannot be found. Instead, a
physical problem during preparation of the film might give a hint of this surprisingly
unexpected result. You observe a larger modulus (larger stiffness) on the 19 mm films
compared to the similar 27 mm ones. This can be due to the used sticks of wood when
making the free films. On the 19 mm wood sticks the paint from the paper adhered in some
degree to the stick itself making it a little difficult to remove. The 27 mm sticks had a slight
different design with bends on the edges making the paper with film easier to remove.
Because of this observation, it is possible that the 19 mm films has an extra thin layer of paint
along the edges making it more stiff (brittle) which can explain the higher modulus on the 19
mm samples. This can be confirmed by looking at this microscope photo in figure 2.

ã 
!""## 
        # 
Comparing the results above with the other paint (45880), only an effect as above mentioned
is seen on the modulus at low speeds. At higher speeds no difference between the two widths
is observed. That the speed should have a larger influence on this particularly paint is difficult
to explain. It is interesting though, that in most of the cases (at speeds higher than 2 mm/min)
you do not observe any difference between the different widths. This could be explained by
the more soft paint. Because of longer curing time, when pulling off the film from the sticks,
it seems easier to remove than the other paint. Even though it is attached in some degree to
the 19 mm stick of wood it does not attach in the same degree as the 15500 paint mentioned
above. The result of this could very well be that the thickness along the edges is not
dominating as much in these films.
ã 
#  "$$#%!





ß
    
 
Before measuring it was expected that the small cracks observed along the edges would
influence the modulus; especially on the 19mm samples. This was however not observed.
Actually the opposite was seen.
As already mentioned above, there is a slight difference in the modulus of the two widths. It
is most pronounced in the brittle paint (15500) because this had a larger degree of attachment
to the 19mm sticks of wood than the other paint had. The 27mm sticks had a different design
making the films easier to remove.

When comparing the different preparation methods it is found that, in case of the 45880 paint
the 27mm stick and the 19mm made using tape are similar. This only contributes to the
theory about the error in the preparation method of the 19mm samples when using sticks.

In order to avoid this in the future, the stick of wood must be approximately 2 mm shorter in
width than the paper placed on top of it. By doing this, paint attached to the stick can be
avoided.

ß
   
 
When comparing the short and long samples, a considerable higher modulus is found for the
long samples (120mm), which also leads to less strain. This is not expected. The tensile
strength is considerably larger for the short samples. This is also not expected.
An explanation for the tensile strength being smaller for the long samples can be, as
mentioned earlier, that the longer the sample, the bigger the chance to have defects within the
sample; especially along the edges.

ß
     
The effect on the modulus from the speed of pulling is believed to have no relations on the
modulus. By neglecting the only result indicating a slight influence of the speed, there is no
relations between the two parameters (see the discussion in "comparison of width", 2nd
section above).
Of some interest, it is found that the paints have breaks at a higher Fmax the higher the pulling
speed. In many cases it is seen that a speed of 50 mm/min results in a larger tensile strength
than the one you get with a speed of 2 mm/min. The strain is about the same.

Example:

?
Comparison of Fmax between the same samples pulled at speed 2 mm/min (up) and 50 mm/min
(down).

?
?

?
V  ? ?

_     _ ?

?? 
??
?? ? ??

??  
? ? 
?


? 
? ?  ? ? 
?  ? ? ?
?  ? ?  ? 

? ? 
?


???? ?
?
 ? ? 
?
?? ? ??? ?
?

? 
? 
?? ? 
?
?

? ? 


? ?
?  ?
? ?  ?  ? 
? ? ?  
?  ?  ?

? ?  ?  ? ? ? 


?  ? ? 
?  ? ? 

?  ? ?  ? ? ?

 ?
? ? ??  ? ?? ? ? ?
?
?
 ??
??? ? ?

?? ?
? ??
? ? ! ???? ?!? ? ?
?"#?$?%&&'?

&()? ?  ? ?  ?  ? *


?  ? 
? &! ? ? ?  ?
? ? ?

 ?
??
?? &!? ??? ? ? ? ?+

?? ?
?

 ? 
? ? 
? ? &! ? ? ?  ? ?  ?  ? ?  ? ?  ?

 ?  ? ? 
? &! ?
? 
? ?  ? ?  ?  ?  ? ?  ? &! ?

? ? ?  ? 


? ? ? 
?  ? ",)? ? ?  ? ? ?


?  ?
? *?  ? ?  ? ?  ? *
?  ? &! ? 
?  ?


?
? ?&! ?
?

-
'?#? 
? ?

? ?  ? ?  ?  ? &! ? 


?  ? !?
?
? ? ? ? ? ? ?
?


 
? ? ? 
?  ?
?  ?  ? 
 ?  
? ?

?  
?

(a) 
atancy correct
on:?
? 
? ? ? 
? ?  ? ? ? ?  ? ?

? ?
?
? ?
?

? ? ? ?
 ??
 ? ??
?

? ?? 


? ?",)?

.
??/ ?"&( )? ??  
?  
?
?? ? ?
?
???

? ??
?,? ?&?

?  ? 


? ? 0000?"&)?

? ?
?? ???
??  ??

? ? ?

(b) erburdenressure orrect


on:?
? ? 
? ? ? ?  ? ? 
? 
 ? ? ?  ? 
?


?  ? 
? 
? ? 
? 

? ? ? ? ? ? 
? ? 
?



? ? 
? ? 
? 
?  ? 1? ? 

? ?
? 
?


?
 ? 
? ?  ? ? 
?  ?  ? 
? ? ?  ?  ?
?


? ? ? ? ?  ?
? 
? - ? 
 
? ? ,? $? ?


?  ? ? 
? ? ? 
?  
? ? ? ?  ?  ? ?

? 
? ? ?

2
??3 .?"&( )? ??? ??  
?*
? ?? ?
? ??

00000"%)?

? ?
? ?  ? ? ?
? ?  ? ? 4 
? ?

? ?

? ?*
?
? 
? ? 5?%6!? ?

?
? ? ?
??!7??%!?? ?
?
???%!?  ??



??%!? ? 
??
???
?

??
?  
? ??
?

?  
? ? ? ?  ? ? 
? ? ? 
? ? ? 

? ?

 
?  ? *
? "%)?
? ? 
?  ? *
? "%)?  ? ? ? ? ?
?

*
?"&)?

?? ?  


?
? 
?
????
 ?  
?
? 
?

Vous aimerez peut-être aussi