Determination of the Surface Roughness of Parts in Hard-to-Reach Areas Using Replicas

G. S. Pushkar, A. A. Asmus

(Submitted by the scientific seminar of the Department of Machine Tools and Metal Cutting and Mechanical-Engineering Technology)

When measuring surface roughness in hard-to-reach places (involute surfaces of gear wheels, front and rear surfaces of cutting tools, holes in parts, etc.), a number of difficulties arise. Therefore, under shop conditions, the quality of a surface is very often assessed by visual comparison with roughness specimens. This method is insufficiently objective and often leads to significant errors. Thus, W. Hagen [3], determining the roughness of the surfaces of the tooth profile of gear wheels, pointed out: “The so-called ‘good appearance’ of the tooth profile, determined by inspection and by feeling with the fingers, often leads to erroneous conclusions, even with great experience. For example, one of the leading firms believed that it was manufacturing wheels with very good surface finish. However, detailed investigations showed that the surface roughness was more than twice as great as that of standard wheels.”

It follows from the example cited that, for a correct assessment of the magnitude of roughness, it is necessary to carry out measurements on special instruments.

In those cases where there is no possibility of making a direct measurement of a portion of the surface of a part, the replica method, investigated in detail by M. G. Golubovskii [1], is widely used.

The essence of the replica method is that a replica is taken from the surface being measured, and the roughness of this replica is subsequently measured on optical or stylus instruments.

The following basic requirements are imposed on the material for making replicas:

1. Accurate copying of the roughness of the measured surface.
2. The replica material must not undergo plastic deformation at the moment of removal or during storage of the replica.
3. The replica material must not “seize” with the surface of the part.

The following are used as materials for making replicas: natural wax, wax mixed with turpentine and graphite, zapon lacquer, celluloid, low-melting alloys, serographite, gelatin, gypsum-chromic peak, oil-gutta-percha compound, and others [1], [2].

Special attention should be given to the use of so-called “transparent replicas.” As a material for making such replicas, ethyl-cellulose, diacetate matted, and nitrocellulose film 0.12–0.16 mm thick may be used. The advantages of transparent replicas are:

the possibility of obtaining good impressions from mechanically machined surfaces;

the possibility of accurately measuring the roughness value of the replica on an interference microscope with the aid of Zehender’s immersion chamber [4];

the possibility of measuring irregularities over a wide range, from 0.1 µm to 30 µm.

At the Department of Mechanical-Engineering Technology, research work was carried out to select a material and refine the technology for producing transparent replicas.

1. Selection of material for making replicas

W. Hagen [3] investigated the possibility of using various materials for making transparent replicas.

It was established that films made of different materials copy surface irregularities with different distortions.

The magnitude of the distortion increases as the roughness height of the specimen increases.

The best results were obtained when using film on a nitrocellulose base. W. Hagen asserts that the most suitable film is one of German manufacture.

In order to refine W. Hagen’s recommendations, we investigated the following materials for making replicas:

basis of X-ray film;
basis of photographic film;
basis of slide film;
basis of triacetate and diacetate motion-picture film;
basis of nitrocellulose motion-picture film.

All the materials were of domestic manufacture.

As a result of the study it was established that the most complete filling of the surface irregularities is achieved by using motion-picture film with a nitrocellulose base.

All the other films soften insufficiently under the action of the solvent and therefore copy the irregularities of the specimen incompletely.

Subsequently, for making replicas, only motion-picture film on a nitrocellulose base was used.

2. Technology for making replicas

The following main factors affect the quality of a replica:

1. The concentration of the solvent applied to the surface of the film or specimen.
2. The method of removing the impression.
3. The drying time of the replica.

Effect of solvent concentration

A solution of nitrocellulose in acetone was used as the solvent. For this purpose, ordinary motion-picture film on a nitrocellulose base was thoroughly washed free of emulsion, degreased with carbon tetrachloride, cut into pieces, and kept in acetone until complete dissolution.

Solvents with a nitrocellulose concentration from 3 to 12% (by weight) were used. Increasing the concentration leads to an increase in the viscosity of the solvent and to a significant content of air bubbles in it, which cannot be removed when the film is applied to the specimen.

A low concentration of nitrocellulose and the use of pure acetone lead either to squeezing out of the solvent during pressing or, with abundant wetting, to softening of the entire volume of the film with a sharp distortion of the height of the irregularities of the specimen.

The highest-quality impressions were obtained when using a solvent with a 4–6% nitrocellulose content.

Method of Taking the Impression

The quality of the impression depends on the method by which the solvent is applied. The following variants were tested:

1. The solvent was applied to the specimen, after which the film was placed on the surface of the specimen (Fig. 1).

Fig. 1.

1. The solvent was applied to the film, followed quickly by placing the film on the specimen.

2. The solvent was applied to the specimen and to the film, after which the film was placed on the specimen.

For specimens with a roughness of no more than 20 µm, it is recommended to use variant 2 for applying the solvent. For specimens with a roughness greater than 20 µm, the third variant of applying the solvent is recommended. This is explained by the fact that when the second variant is used, the softened layer of the film does not completely fill large depressions of the irregularities. When the solvent is applied to the specimen, more complete filling of the irregularities occurs, and a more accurate copy of the specimen irregularities is formed on the replica.

After the film is placed on the specimen, light pressing is performed in order to remove air bubbles and excess solvent between the film and the specimen.

The duration of pressing is 1–2 minutes.

The following were tested:

pressing by applying pressure with the fingers;

pressing with rubber and a plate;

pressing with a plate having a lapped surface.

Pressing with the fingers and with rubber produces waviness on the base surface, which leads to a substantial reduction in the roughness of the replica in comparison with the roughness of the specimen.

Pressing with a plate gives the most even reference surface. It is desirable, during pressing, to move the plate in a direction perpendicular to the direction of the irregularities.

Effect of the Drying Time of Replicas

According to V. Hagen [3], drying of a film wetted with solvent lasts approximately 5 minutes, after which the film in most cases springs back.

According to M. G. Golubovskii [1], the drying time ranges from 2 to 10 min for smooth surfaces and from 5 to 40 min for rough surfaces.

According to the results we obtained, for a 5% solution of nitrocellulose in acetone and a drying temperature of 18–20°C, the drying time is 20–40 min, which agrees with the data of M. G. Golubovskii. It should be noted that, on insufficiently dried replicas, because of their plastic deformation, an increase in roughness is observed in comparison with the roughness of the specimen.

Analysis of the Causes of Distortion in the Magnitude of Replica Irregularities

According to various sources, the magnitude of the irregularities of a replica is obtained as lower than the irregularities of the specimen.

Thus, V. Hagen [3] indicates that, with a roughness of more than 30 µm, the reduction in the height of the replica irregularities is 10%.

According to M. O. Jakobson [2], the distortion of irregularities for various degrees of smoothness is 5–7%.

Neither V. Hagen nor M. O. Jakobson indicates the causes that bring about a reduction of the irregularities on the replica.

In order to determine the causes of distortion of the replica irregularities, special work was carried out.

A nitrocellulose motion-picture film, freed of the photosensitive emulsion, was coated with solvent (a 5% solution of nitrocellulose in acetone) and placed on the specimen, pressed with a lapped plate, after which the specimen together with the film was set in the field of view of an MIS-11 microscope.

During the drying period of the replica (up to 30 min), the nature of the film deformation was observed under the microscope.

It was established that, when the height of the specimen irregularities was up to 20 µm, deformation of the film was practically absent.

Subsequent measurements of the height of the irregularities of the specimen and of the replica showed good agreement between the magnitudes of the irregularities on the specimen and on the replica (Table 2).

When the height of the surface irregularities of the specimen was from 20 to 60 µm, it was clearly visible under the microscope that, 8–10 min after the start of drying, the upper reference surface of the film began to deform (Fig. 2).

The magnitude of deformation “a” reaches 30–50% of the height of the irregularities.

After removal of the replica, some additional deformation of the film occurs; the magnitude “a” decreases, and the height of the replica irregularities is distorted by the same amount.

This phenomenon can explain the reduction of the irregularities on the replica that was noted in the works of V. Hagen and M. O. Jakobson.

In order to reduce distortion of the height of the replica irregularities, the technological process for producing replicas when measuring irregularities of the order of 20–60 µm was changed.

The following technological process for making replicas was finally adopted:

1. Wash off the photosensitive layer of the nitrocellulose film with hot water.

Table 1

Quality of copying roughnesses with a height of more than 20 µm by single- and double-film replicas

1. Dry the film.
2. Cut the film into pieces measuring \(10 \times 30\) mm.
3. Thoroughly degrease the surface of the film and of the specimen with carbon tetrachloride (analytical grade).

Film
Specimen

a) beginning of drying;
b) end of drying.

Fig. 2.

1. Dry for 1–2 min.
2. Coat the film with a 5% solution of nitrocellulose in acetone (analytical grade).
3. Place the film on the specimen and press it with a ground plate in order to squeeze out air bubbles.
4. Dry for 30 minutes.
5. Wipe the upper surface of the film and the surface of the second film with carbon tetrachloride.
6. Apply a layer of solvent to the surface of the second film and press.
7. Dry for 20–30 min. After this the replica is ready for measurement.

Replicas made by the new technological process (two-layer) were observed under an MIS-11 microscope throughout the entire drying period.

The amount of deformation in the manufacture of two-layer replicas is considerably smaller than in single-layer replicas (Fig. 3).

a) deformation of a single-layer replica,
b) deformation of a two-layer replica.

Fig. 3.

For subsequent measurements it is necessary either to glue the replica, with its base surface, onto a smooth plate, or to fasten it in a special fixture.

We carried out comparative measurements of irregularities using single-layer and two-layer replicas (see Table 1).

Table 2

Quality of copying roughnesses with heights of less than 25 µm by single-film replicas

From Table 1 it is clearly seen that, when two-layer replicas are used, the distortion of the irregularities is smaller than when single-layer replicas are used.

Further experiments were subsequently carried out to establish the possibility of copying surface irregularities using two-layer replicas.

Table 3

Quality of copying of irregularities by replicas made from nitrocellulose film (two-layer replicas)

Irregularities with a height from 17 to 60 µ were measured. The results of the measurements are given in Table 3, from which it is evident that the distortion in the magnitude of the irregularities does not exceed the measurement error on optical instruments.

Conclusions

1. Ordinary nitrocellulose-based photographic film may be used as the material for making replicas.

2. As the solvent, a solution of nitrocellulose in acetone should be used. The optimal concentration of the solution is 4–6% nitrocellulose (by weight).

3. When measuring irregularities of the order of 20–60 µ, two-layer replicas must be made according to the technology described above.

References

1. Golubovskii M. G. Replica method for checking surface finish. In: Quality of the Surface of Machine Parts, LONITOMASH, 1949.

2. Yakobson M. O. Roughness, work hardening, and residual stresses in machining. Mashgiz, 1956.

3. Hagen W. Der Einfluß von Werkstoff, Maschine und Werkzeug auf die Oberflächengüte der Zahnflanken. VDI—Z. Bd 97, Nr. 25; 27, 1955.

4. Zehender E. Ein Interferenzverfahren zur Untersuchung rauher Oberflächen. VDI—Z. Bd 94, Nr. 14/15, 1952.