Set Up of Mar-Scratch test Methods

Author:European Union Publications Office, 2006

1. Activities and Results at Voestalpine. 1.1. State of the art in scratch-resistance and scratch testing. 1.2. Evaluation of technological scratch tests at Voestalpine. 1.3. Evaluation of physical properties related to scratch resistance. 2. Activities and Results at Arcelor. 2.1. Scratch test and reference materials resistance. 3. Activities and Results at Doc. 3.1. Scratch and abrasion test. 3.2. Indents and micro scratches. 4. Activities and Results at CSM. 4.1. Scratch test. 4.2. Dry-scrub resistance test. 4.3. Nail - mar test. 4.4. Nano scratches (Berkovitch indenter).


The analysis of existing testing methods has been started from a bibliographic examination of the work done in this field.

In recent years the study of mar and scratch resistance of painted products has registered a growing interest especially due to the increasing needs of car manufacturers, requiring better aesthetical quality and durability of the automotive topcoats. In the aim of developing coatings with enhanced performance the areas that have received more attention by researchers are the study of the nature of the defects and the definition of laboratory testing methods.

Many efforts have been done in order to correlate the mechanical properties of the coatings with their resistance to aesthetical damage. With this purpose, various coating characteristics have been showed to have an high influence on mar resistance: hardness, yield stress, elongation, elastic modulus, etc. ...

Contributes to literature review are given in the following paragraph and in the 1st Report.

In literature a widely accepted classification of damages concerns their depth (see figure 2-1): damages with a little penetration respect to the total coating thickness (e.g. 0.1-2 um) are usually referred as "mar", while the more common term of "scratch" is better referred to deeper damages (e.g. >2 um).


The examination of mars and scratches shows that different morphologies can be identified; the most common classification is between damages consisting in plastic flow or in rupture of the coating (figure 2-2).

The opinion that coating ruptures are more easily seen by the observer is sustained by some authors.


Various types of test methods exist in order to evaluate the material resistance to scratches. Many of them already existing have been evaluated in the course of the project; some have been adapted, focusing on the kinds of new coatings, to measure the aesthetical damages.

1. Activities and Results at Voestalpine

1.1. State of the art in scratch-resistance and scratch testing

In a first step a literature review was performed in order to characterise different types of scratches that occur on coated surfaces. Focusing on the polymer network two types of deformation can be defined according to literature [1,2,3]: plastic deformation and brittle (= fractioned) deformation. A difference in optical appearance between these types is reported.

Plastic deformation

With plastic deformation the surface of the coating is changed, but the connections of the network structure remains unaffected. If the deformation is a scratch type one it can be characterised as a smooth groove attended by a shoulder on each side as shown in the upper part of figure 2-3. Along the direction of the scratch no discontinuities occur, so that the reflection of light does not significantly differ to that of the unaffected surface. In the perpendicular direction, the reflective angle is modified by the scratch morphology. The visual appearance of scratches with plastic deformation therefore depends on the direction they are looked upon. In longitudinal view they may even remain unrecognised. If gloss is measured, the degree of reflection depends on the orientation of the scratches. Scratches with plastic deformation may be healed up by plastic flow under the application of heat.


Brittle deformation

With brittle deformation the structure of the polymer network is broken. Scratches loose their continuity in longitudinal direction as cracking occurs. The scratches may even reach the metal surface. The diffraction of light is now independent from the incident direction as is shown in the lower part of figure 2-3. Parts of the coating may be lost, in severe cases this effect is called abrasion. Brittle deformation of scratches is not reversible.

It is obvious that the ability of a coating to undergo plastic deformation depends on its elastic properties. E' - the elastic response or storage modulus - and E'' - the viscous response - are the main parameters to describe this behaviour. The ability of a coating to perform plastic deformation depends on E' as well as Tg - the glass transition temperature - [4]. Below Tg , E' strongly depends on temperature. Whereas above Tg E' is nearly independent from temperature. E' and tan- (= E''/E') - the ratio of viscous to elastic response - are measured by DMA (dynamic mechanic analysis). The maximum tan- indicates Tg in the DMA plot. With E'min the crosslink density ue of a polymer can be calculated according to ue = E'min / RT under favourable conditions. ue gives information on the completeness of curing of a polymer.

A high degree in hardness will result in brittle deformation thus leading to the worst (i.e. irreversible) case of marring. Correlations between mar resistance and hardness have been found as well as experimental evidence for the influence of E' on the scratching behaviour [5].

Chemical resistance and resistance to diffusion usually correlate to the hardness of a polymer as well. Consequently the mar resistance of coatings is adverse to its chemical resistance [6,7,8]. Future investigations should aim to find coatings that combine a high elastic modulus with high chemical and diffusion resistance.

Summing up it can be said that physical testing (tan-, E', E'', universal hardness ...) is a useful supplement to technological scratch tests because it allows the establishment of material laws and of damage mechanisms. The application of both tools is a pre-requisite for a more straightforward coating development.

1.2. Evaluation of technological scratch tests at Voestalpine

The scratch tests chosen at voestalpine (blue underlaied in figure 2-4) cover the whole spectrum of damages that can occur due to abrasive degradation of polymer surfaces. A closer description of these tests and the appropriate analyses is summed up in table 2-1.

Table 2-1: Testing Methods, Sample Sizes, Analyses and Standards for the Investigations on mar resistance of coated panels performed by VASL.

Method Sample Size [mm] Analysis Standard
Erichsen Single Scratch (dry) 90 x 90 * Spalling at the Edges (Photographic Documentation)
* Lasertopography
According to ISO 4586-2:1997 (E)
Amtec (wet) 100 x 250 * Loss of Gloss before and after Thermal Curing DIN 55668 Entwurf  (September 2000)
Taber (dry) 90 x 90 * Lasertopography
* Loss of Gloss before and after Thermal Curing
ASTM D 4060 - 95
Linear Scrub Tester +  Commercial  Cleaning Emulsion (wet) 100 x 300 * Loss of Gloss before and after Thermal Curing DIN 53778 Part 2 (August 1983)
Falling Sand Test (dry) 100 x 150 * Lasertopography ASTM D 968 – 81 (Reapproved 1986) With Modifications


Scrub - and brush type tests

Mar resistance is of major interest for the automotive industry but also for the front parts of coated domestic appliance goods. Marring is caused by a complex pattern of scratches typically with a depth of - 2um. This marring-process is simulated by scrub- and brush type tests like the Amtec-Test or by Linear Scrub Devices both in combination with a variety of abrasives.

As the scratches have various directions, and therefore often intersect, the surface looks dull after treatment. Evaluation of the samples is done by gloss measurements before and after testing. A representative examination of single scratches is impossible but the loss of gloss is a good measure for mar resistance.

The effect of healing scratches by heat can also be measured by the gain in gloss after heat treatment. This "reflow effect" is an important aspect in the development of new coating systems, standard reflow conditions used in this work are 2h at 80°C.


This test has been developed to determine the mar resistance of automotive clearcoats ("laboratory car wash equipment"), manufacturer of the testing device is Amtec Kistler GmbH in Germany. The testing procedure is described in literature [5], testing principle is the scratching of a clearcoat surface with a rotating brush (figure 2-5) and a silica containing aqueous slurry. Method development at voestalpine revealed that tested samples are covered by a dull polyethylene (PE) layer resulting from residuals of the rotating brush (figure 2-6). Thus the cleaning procedure of the tested samples is of great importance. Rinsed specimen which are still covered by PE exhibit a strong gloss loss in comparison to the wiped state where this film is rubbed off with a piece of cloth. So far this cleaning procedure has not been specified exactly in the proposal of the standard used for this testing method (E DIN 55668: 2000-9). Wiping off the PE-residuals has been used as a standard cleaning procedure.

Standard conditions used throughout this work are as follows:

- Abrasive: 1,5 g/l Sikron SH200 in water

- Motion of the test table: 10 double strokes at 5m/min



Stylus testing

For detailed information on elasticity and hardness of a coating it is necessary to gain information on the depth of a deformation caused by a tip in correlation with the applied normal force on this tip [1,9,10]. Scrub - and brush type tests do not give detailed information on the force of a single particle of the abrasive being pressed upon the surface. In addition the scratches also might intersect and so no clear statement can be achieved with those tests. Here stylus testing is the method of choice.

Erichsen Single Scratch Test

This test is a modification of a stylus scratch test as described in ISO 4586-2:1997 (E). A diamond tip is pressed upon the surface with increasing normal force while the panel is moved forward underneath the tip. The testing principle is shown in figure 2-7, improvements have been made in respect of working range, resolution and especially evaluation techniques. With low loads typically plastic deformation occurs, rising loads result in brittle deformation with an increasing number of cracks. Optical microscopy of scratched samples (figure 2-8) is used to measure scratch width, to describe the cracking behaviour of the coated samples and the "Critical Load" i.e. the test load where the first crack is formed. Further information on scratch depth and pile up height can be obtained by laser profilometry: an xy-grid of heights on a surface can be gained at an accuracy of aproxx. +/- 10 nm without affecting the topography of the sample.


Severe abrasion tests

For testing the resistance to abrasion - i.e. a significant loss of material - the Taber Abraser and the Falling Sand Test are used.

Falling Sand Test

From the apparatus shown in figure 2-9, sand of defined quality is falling onto the substrate at an angle of 45°.

The standard ATSM D 968-81 suggests to measure the amount of abrasive needed until the first part of the metal surface lies blank. This method has been modified as follows: only a small rim of 5 mm width lays blank and the rest of the surface is covered with adhesive tape. The covered (unaffected) area is used as a reference. The maximum depth of abrasion after certain amounts of abrasive (normally 5 liter per "step") is detected via lasertopography and the maximum depth is indicated at each profile. With our reference material "CH26" an excellent linear correlation is found between the amount of abrasive and maximum depth of abrasion (figure 2-10). A volume of 20l sand turned out to be a significant quantity for singular testing.

Among the test methods with severe abrasive conditions the "Falling Sand Test" turned out to provide more reliable test results in comparison to the "Taber Test" : the latter suffers from a low weight loss under the experimental conditions chosen and thus a poor reproducibility and differentiation between different clearcoat types is obtained (figure 2-11). For this reason the Taber test has only been used in the starting phase of the project.



1.3. Evaluation of physical properties related to scratch resistance

Universal Hardness (Micro-scale instrumented indentation)

J.L. Courter [11] reported that high mar resistance in a dry abrasion test is associated with: - high penetration depth under load (low universal hardness, HU, hardness under load), - low indentation modulus,

- high total indentation work, high residual indentation depth after unloading (which would correspond to a low traditional hardness, such as Knoop hardness, where the larger the residual indentation, the lower the hardness), and

- high creep rate.

Hardness measurements have been performed on a Fischerscope H100 S with the following instrument settings:

F max = 10 mN

Load cycle: 20 sec. - dwell time: 30 sec. - unload cycle: 20 sec. - dwell time: 30 sec.

Dynamic viscoelasticity measurements

In literature [12] a relationship is reported between the storage modulus G' of clearcoats and the ratio of the scratched area obtained from a wet marring-test. Preliminary measurements of voestalpine on free clearcoat films at 40°C and 55% r.h on a Paar Physica US 200 Rheometer (parallel plate system) suffered from poor reproducibility (figure 2-12).


The testing method has been improved within this project substantially due to the following measures: - in-situ-crosslinking of the clearcoat: improves adhesion to the oscillating plates of the US 200 and results in a better load transmission,

- correction of the film thickness: additional measurements of the film thickness after each test cycle where necessary for the thickness reading of the US 200 was not exact,

- determination of the working range (i.e. range of linear viscoelasticity) with an Amplitude-Sweep-Experiment (figure 2-13).


2. Activities and Results at Arcelor

Mechanical abrasion of stainless steel during use, and its consequences on the steels properties depends on the application. In household appliances, one of the most important markets for stainless steel, scratches can be made by a knife when you cur food on a working surface or by scotch brite pad with abrasive powder when you clean the surface. These scratches can damage different properties of stainless steel such as the resistance to corrosion, food compatibility, hygienic or cosmetic aspects.

2.1. Scratch test and reference materials resistance

The Clemen test (figure 2-14) is classically used to qualify the scratch resistance of organic coatings. This test could be interesting for all our surfaces in an aesthetic point of view. The indent is 1mm CW roll and the measurement unit is gram.

Scratch resistance is visually evaluated. As far as Clemen test is concerned, the load, when scratch is visible, is given.


Results obtained with the reference materials (see table 1-1) are not so good. We cannot notice any distinction between the samples. The roughness of brushed surface slightly hides the damages but the load for which the scratch become visible do not exceed 100g. This is relatively low. These materials do not resist well to scratch test.

3. Activities and Results at Doc

3.1. Scratch and abrasion test


With the aim to create a test which is close to the stresses on the surfaces in the household, at DOC a new scratch and abrasion test had been designed.

The figure 2-15 shows the principle of this test: a stress medium was fixed on a metal tube with a contact area of 2.5 cm2 by a hose clamp. Then the gauze was rubbed over the surface in a trace of 100 mm length for a defined number of double strokes. One double stroke corresponds to a forth and back movement of the metal tube. The tube was moved with the aid of a flexible hose in order to prevent the examinant from exerting an additional force on the metal tube during the scratch test.


Two different stress media were tried.

The first one was gauze soaked with 0.2 ml of "Stahlfix classic £

", a typical stainless steel cleaner. Additionally, 0.8 ml of the cleaner was directly poured onto the area which was to be tested. "Stahlfix classic" contains aluminium oxide as abrasive medium in a broad distribution of particle sizes. The maximum number of particles has a size between 1 and 3 um. Additionally, a smaller fraction in the region between 10 um and 100 um was found. The size distribution function of the abrasive particles is given in figure 2-16.

The second stress medium was a "Scotch Brite £ " cloth which was soaked by "Stahlfix matt £ ", a stainless steel cleaner without any abrasives.

The changes in the surface which were generated by the scratch traces were quantified by means of gloss and AFM measurements as a function of the number of double strokes.

The damages caused by the scratch and abrasion test were evaluated according to the standard EN 13423-2 by the measurement of the gloss values before and after the test. The angle of incidence was 85° chosen since the generally highly reflective stainless steel surfaces cause an overload in the reflectometer if geometries with smaller angles of incidence were chosen.

The more damages occurred the higher were the gloss changes after the scratch test. Both a roughening of the surfaces resulting in decreased gloss and a polishing of the surfaces resulting in increased gloss were observed.


The figure 2-17 represents the progression of the gloss values of the uncoated stainless steel surfaces during a scratch and abrasion test using "Stahlfix classic £". Before starting the scratch test the bright surfaces exhibited, as to be expected, the highest gloss values. During the scratch test, the gloss of the bright 1.4016 decreases, whereas the gloss of the 1.4301 surfaces increases. The reason for the latter observations might be a polishing effect due to the large amount of abrasive particles with sizes between 1 and 3 um.

The scratch and abrasion test with the two different scratch media was applied on the Senocoil coated surface, too. In figure 2-18 the gloss values are compared which were determined on Senocoil coated surface after scratch tests.


The initial gloss value of the unscratched surface is a little bit lower than in the case of the uncoated substrate material. The gloss changes after the test with "Stahlfix classic" were higher than in the case of the uncoated substrate material. Especially the large decrease after only 10 double strokes proves that this coating exhibits a low scratch resistance. The damages which were generated by "Scotch Brite" on the Senocoil coated surface were even higher than in the case of "Stahlfix classic". After 50 double strokes using "Scotch brite" as scratch medium the coating was completely destroyed. The Senocoil surfaces scratched by "Stahlfix classic £" (10 and 50 double strokes, respectively) were mapped by means of AFM (figure 2-19). In contrast to the smooth unstressed surface (see figure 1-5), a strong roughness and waviness was observed after scratching. After 50 double strokes a coating thickness reduction of 2 um was determined by AFM. So almost the complete coating had been removed.


The damages with the second stress medium ("Scotch Brite £" cloth which was soaked by "Stahlfix matt) were too strong to obtain a satisfactory differentiation between the different coatings (figure 2- 18). For this reason, the scratch and abrasion test was preferentially performed with the gauze soaked with "Stahlfix classic" as stress medium.


In order to test the reproducibility of the scratch test developed at DOC, the test was performed up to 50 times on different surfaces:

- an uncoated bright surface of 1.4016

- the coating "Senocoil" applied on brushed 1.4301

- coil-coated carbon steel with additional clear coats delivered by VASL (DC 05 ZE-P 75/75 + 15 um Primer (PU) + 15 um black Basecoat (PU) + 15 um Clearcoat A (PU) + 15 um Clearcoat B (Flour / PU)); sample code: 440389-2245.


The figure 2-20 represents the mean values of gloss and residual gloss after the scratch and abrasion test on the three different materials. Whereas the gloss of the coil coating material decreases continuously, the bright 1.4016 and the Senocoil surface show only at the beginning decreasing and later due to the polishing effect increasing glosses. The two coated surfaces show error bars for the absolute gloss values with a maximum length of ± 1.2 units. In the case of the uncoated substrate, the scratch test is, as to be expected, better reproducible with maximum error bars for the absolute gloss values of ± 0.3 degrees. For the residual gloss values the maximum relative error are ± 1.6 % for the coil coating material, ± 1.7 % for Senocoil and ± 0.4 % for the uncoated stainless steel. Thus, after sufficient measurements on different places of a surface, the mean values can be determined very exactly. In the figures 2-21 ÷2-23 the amounts of different gloss measurement results after the scratch test on the three different surfaces are shown. The diagram shows that with increasing number of double strokes, the distribution of gloss results becomes broader. This effect is stronger for the coated surfaces than for the uncoated one. Thus, in order to get reliable results it was decided that for each material the scratch test has to be repeated at least three times.


3.2. Indents and micro scratches

The surface hardness and flexibility was tested by using the needle of an AFM. Two different procedures were followed. On the one hand indents with the shape of a triangle pyramid were pressed into the surface using a load of 42 ug which was generated by applying a voltage of 1.0 V to the needle. Afterwards, the indent depths referring to the sample surface was measured by AFM. On the other hand scratches of 2 um length were prepared by the AFM needle using different loads between 4.2 ug (voltage: 0.1 V) and 42 ug (voltage: 1 V). The average depth of the scratch with respect to the plane of sample surface was measured again using the AFM. In figure 2-24, examples for the resulting surfaces after indentations and a scratch, respectively, are shown.

In both procedures the applied loads are very small compared to standard tests such as the pencil hardness. The reason for using such small loads was that the aimed coating thickness for clear coats on stainless steel is around 3 um in order maintain the impression of a stainless steel surface.


On such thin coatings the indentation depths should not exceed 1/10 of the coating thickness in order not to include influence of the hardness of the substrate in the measurement, but to determine the hardness of the coating only. During the single scratch test material is removed from the scratch trace, and thus, the damage cannot heal, but is irreversible. Thus, the depth of the scratch characterises the hardness of the coating. After indentation on the one hand the depth also decreases with higher hardness of the coating. On the other hand, however, since material is not removed the coating has the possibility to partly reflow at places where the network of the polymer structure was deformed only and no bond breakage occurred. For these reasons the depth of the indents depends on both, hardness and flexibility of the considered coating.

4. Activities and Results at CSM

4.1. Scratch test

In order to explore the scratch behaviour of the coatings in a wide range of forces, a scratch hardness tester has been used with the various reference materials. In this test a stylus is moved on the surface of the sample, producing a scratch with a force that can be selected in the range 1-20 N, with resolution 1 N. A carbide tip stylus with 1 mm tip diameter (Clemen stylus) has been used (figure 2-25).

Even if this test has been conceived to test only the coatings (and classify them on the basis of the maximum scratching load they can resist without discovering the substrate), in this case it has been used also with uncoated materials in order to quantify the protective effect of the coatings.


On the various reference samples (see table 1-1) parallel scratches have been realised with the following different loads: 1, 2, 3, 4, 5, 6, 8, 10, 15, 20 N. The scratches have been examined by means of optical and electronic microscopes in order to evaluate their morphology and width. Examples of images taken at the electron microscope are given in the figure 2-26 ÷ figure 2-28.

Coated products show a different behaviour respect to uncoated: at the higher loads the coatings are completely broken by the stylus; the width of the scratch on the coating and on the substrate have been separately measured.

The figure 2-29 shows the average width of the scratches obtained at the different load values, for coated or uncoated materials; in the graph the width of the scratch on the coatings has been plotted with dashed lines, while the continuous lines have been used to indicate the scratches on the base metal. For the coated samples the minimum load bringing to complete coating rupture has also been indicated.

Results indicate that: - the use of clear coatings on stainless steel is effective in reducing the scratch width on the base steel, especially at lower load values,

- this protective effect at low loads is higher with the thick plastic film (sample n. 5) than with the thin layer of clear paint (sample n. 4),

- when the load exceeds the coating resistance, there is a sudden increase in the scratch width on the polymeric layer,

- at high loads the width of the scratch on the coating is higher with a thick plastic film than with a thin layer of clear paint,

- some differences can be detected in the behaviour of the different uncoated stainless steel, but, to understand if they are significant, they should probably be investigated with different techniques.

This test method associated with scanning electron microscope is useful to evaluate the behaviour of the coated material at relatively high loads: scratches can damage properties of coated steels such as resistance to corrosion, food compatibility, hygienic aspects. From the aesthetically point of view, however, on all samples damages resulted visible already at 1 N. The sensibility of the instrument at CSM was therefore to low for evaluating lower loads (mar loads).




4.2. Dry-scrub resistance test

To evaluate the dry-scrub resistance of the reference materials and to verify the possibility to improve the performances by applying UV varnishes, it has been used a scrub tester apparatus according to ISO 11998 (figure 2-30).

The test consists in scrubbing the surface of the test panels with abrasive pads of non-woven plastic material ("3M Scotch Brite..." hand pads kinds, size: 90 x 39 mm). Respect to this standard, the procedure has been modified as nor washing liquids or soiling agents have been utilised.

This test has been chosen thinking about possible uses, where the resistance to accidental but numerous scrubbings could be an appreciate property (e.g.: corridors, escalators, etc.).


First tests regarded reference materials and have been performed by scrubbing the surfaces (200 double stroke, length 100 mm) both in the rolling direction and in the cross one. The load was 3.9 g/cm2. The damages have been examined by:

- scanning electron microscope exams

- measurements of colour and gloss variations

- visual observations.

The degree of the mechanical damages observed with SEM are shown in the figure 2-31 ÷ figure 2-36.

The micrographs evidence that:

- the surfaces seem to be more damaged when the scrubs are made in the rolling direction with this kind of exams a certain differentiation among uncoated materials is given, whilst as regards organic coated samples, it must be taken in account the differences in coating thickness.



The ranking damages comparison among the different materials made by visual exams resulted difficult because various factors must be taken into account:

- kind of surface (steel finishing, coated or uncoated)

- lighting orientation

- rubbing direction respect to rolling direction

- samples orientation (vertical / horizontal panels and rolling/ cross direction) respect to the observer.

In spite of all these factors, a visual average ranking has been made, with a good reproducibility obtained among different operators.

A summary comparison of damages evaluation produced by dry-scrub test is given in the table 2-2.

From this table we can say that:

- the ranking by visual observations exhibit some relationships with the colour variations (average values in the cross and rolling directions)

- the ranking by SEM exams seems to be related more to the depth of the mechanical damages (uncoated steel sheets) and/or to the removing of the coatings

- the gloss variations for high gloss surfaces (>100) could be too much indeterminate, however from the table it seems that values obtained on scrubs in the cross direction have some correspondences with the damage degree observed by SEM exams.

Table 2-2: Comparison of damages evaluations

  Samples Damages Visual appearance Colour E Damages SEM appearance Gloss (85°)  %
          rolling cross
1 1.4301 BA +++++ 2.6 + 5 5
2 1.4301 brushed +++++ 2.5 + 16 4
3 1.4016 BA +++++ 3.7 ++ 9 13
4 1.4301  brushed  + clear coat ++++ 1.0 +++ 22 32
5 1.4016 brushed  + plastic film +++ 0.6 ++ 15 10
6 prepainted HDG ++++ 1.6 ++ 25 14

Note: + less damaged, +++++ more damaged

During the course of the Project further scrubbing conditions were proved by using also less abrasive scrubbing pads and varying the number of double strokes (table 2-3).

Table 2-3: Description of the scrub resistance tests performed

Conditions (for all tests)  size of scrubbing pads:              90 x 39 mm
        load:        ∼4 g/cm2
        scrubbing of the surfaces:  in the cross direction
Evaluations of damages by visual observations
  1st kind 2nd kind 3rd kind
scrubbing pads “3M  Scotch  Brite classic” “3M  Scotch  Brite bagno” (abrasive side) non-abrasive sponge
code RV 750313827 RV 75031892 ---
N° of double strokes 200 10 and 25 200
wet/dry dry dry wet  (1 ml H2O+1.5 ml of detergent)

From the tests made on new samples, some good results were found. In particular a new coating (an UV curable one) had a high scrubbing resistance even in the more severe conditions (200 double strokes with "3M Scotch Brite... classic" pads). The same material exhibited in the 3rd condition (less severe) damages hardly visible with the naked eye.

Generally quite the same ranking among materials were observed with different scrubbing conditions. The best one to compare the wider range of coated materials was the second condition (with medium abrasive pads).

The better visual observation conditions to evaluate different materials resulted as follows: - diffuse light

- samples in horizontal (flat) position

- angle of observation -20° from the surface perpendicular


4.3. Nail - mar test

The resistance to nail scar damages has been evaluated with the Mar Tester (Erichsen Model 435). This instrument, figure 2-38, consists in a scarring tool in the form of a disc, with locking facility and made of special plastic or metal, mounted on a screw and under pressure from a helical spring. The marking disc is put between two rotating guide wheels. The force is adjustable from 0 to 20N. The instrument is placed perpendicularly onto the test surface, pressed down and moved for few cm. The result is the spring force which produces a scar visible with the naked eye, but not a scratch. The reproducibility of this test is higher than the traditional test consisting in trying to mark the surface with a fingernail. Moreover the evaluations are simple. During the tests, made with the plastic disc, it was observed on some samples that the scar is not visible any more after an elapsing time variable from 15 to 30 minutes. A relaxing effect due to temperature (few minutes at -60 C°) was also observed with the same samples. Therefore new coated samples have been evaluated 1 hour after having made the scar.


4.4. Nano scratches (Berkovitch indenter)

The relaxing effect time-dependent was investigated applying nano-scratch with a Berkovitch indenter on some materials selected for their different nature.

The samples were (see also table 5-28):

- B/05/03, coat HC 63, thermally curable

- F/04/03, coat N 1000, UV curable

- C/05/03, coat UV 39M, UV curable varnish applied on a thermal cured primer.

Among the above samples, the ones with code B/05/03 exhibited relaxing effect both with time and with temperature with the Nail-Mar tester.

A set of observations with AFM microscope have been made immediately after making the scratch and then at different times. Scratch tests conditions are shown in the table 2-4.


From AFM topography observations it results an evolution of the scratches with time for two of the three materials tested. To describe this effect, the variations of roughness "Ra" have been calculated and reported in figure 2-39. It can be seen that for the sample F/04/03 (UV coat) no relaxing effect occurs, and therefore the roughness values remain constant. The behaviour of the B/05/03 sample (thermally cured) is very different with a reduction of "Ra" with time till to a disappearing of the visibility of the scratch. Examples of the AFM observations during the time relative to the sample B/05/03 are shown in the figure 2-40.

An intermediate behaviour (i.e. a slower "Ra" reduction) has been observed on the C/05/03 sample (UV varnish, applied on pre-primed surface).