Chapter 3 – Enhancing the users understanding of abrasives and grinding wheels

 

As set out in Section 2.1, the objectives of the study are to facilitate the end user to do the following:

1. Understanding the effect of different grain types and forms
2. Learning to assess safety of grinding wheels
3. Learning to characterise a grinding wheel and
4. Understanding equivalent gradings between different suppliers’ wheels

3.1.  Understanding the effect of different grain types and forms

3.1.1. The objective:

Grinding is an important process in most metal removal applications. Even though many types of abrasives are in use, the most versatile one is alumina abrasive. Alumina is available in many varieties based on the impurity levels of other components like silica, iron oxide, calcia and titania. Of these, titania is known to impart toughness to the abrasive. Based on the amount of titania present, alumina is of two types- the regular and the semi friable variety.

Any grain can be used in its normal form or in a processed form as agglomerates. Grits of abrasives are first processed using vitrified or resin bonds and agglomerated to spherical shapes. These are strengthened by curing for resin bonds and by vitrifying the ceramic bonds. In order to substitute a particular application by an agglomerated product, the final physical size of the agglomerate has to be more or less equal to the single grit being used for the application. Otherwise, it may drastically change the cutting dynamics due to large initial cutting action due to larger grit size of agglomeration.

In products using such agglomerated grains, there are two different types of bonds that influence the wear of the abrasive. One is the bond that holds the grains in the agglomerate and the other which holds the agglomerates to the wheel.
Grinding wheels are either manually operated as in depressed centre (DC) wheel application or machine operated as in cylindrical grinding.

The purpose of this study is to evaluate

• the effect of varying content of titania on the abrasive performance of alumina in both the manual and machine operations; and
• the effect of agglomeration on the performance of a given abrasive.

 3.1.2. Experimental plan:

Two types of alumina abrasives with differing titania contents were chosen for the study. The properties are compared in Table 3.1. The one with around 2-3% titania is generally called as regular brown alumina (A) and the other one with less than 1.75% titania is generally called as semi friable alumina (SA).

For agglomerated grains, wheels were made with agglomerates made out of vitrified bond as well as organic bond. The wheel was made with organic bond.

 

Table 3.1 Chemistry of brown and semifriable alumina
Oxides	A	SA
Al2O3	95.5	97
TiO2	2.7	1.3
Fe2O3	0.2	0.16
SiO2	0.8	0.53
CaO	0.8	0.51

 

These grains were fabricated into abrasive discs, abrasive belts and grinding wheels. The performance of these abrasive products were tested using the standard testing procedures followed in the Grinding Lab and Application Development Centre of Carborundum Universal Ltd. The details of the testing covering the manual and machine applications are listed in Table 3.2.

Table 3.2 Machine details and the cutting conditions

 

a) For Manual wheel evaluation of DC wheels (Depressed Centre)  Bosch make Wheel Size: 7" Workpiece: Mildsteel Time of test: 30 mts

b) For Cylindrical wheel evaluation Universal Type Workpiece Size: 100x10x70mm Workpiece: MS Time of test: 10mts Wheel Size: 300x 13 x 127 mm Wheel rpm: 1710 Work rpm: 234 Feedrate: 4mm/min Total Stock: 8mm/4 minute

c) For Coated Belt test
Machine	Auto Belt Grinding Machine - Horizontal Type
Work piece	Mild steel flat- Size: 10 cm width; 2 m long
Time of test	10 minutes
Parameters monitored	Wear, Metal removal rate, GR,Stretch, Finish
Size of the belt used	50 x 2000 mm.

 

The results were tabulated in terms of the grinding ratio (GR) and metal removal rate (MRR) in Table 3.3. , Table 3.4.  and Table 3.5. The formula used are:

GR =metal removed in grams/ abrasive wear in grams
MRR = metal removed in grams / time of grinding in minutes

3.1.3. Observations and discussions:

i) The effect of grain type on performance is tabulated in Table 3.3:

Table 3.3 Effect of grain type on grinding performance

Application	Grain type	MRR gm/min	GR	Remarks
DC wheels	A	28.5	4.6	GR higher by 35%
	SA	25.3	6.2
Cylindrical wheel	A	5.7	54.46	GR higher by 13%
	SA	5.8	61.68

 

In all the applications, Semi friable grains (SA) showed improvement in the life of the wheel.

ii) The effect of agglomeration on performance is shown in Table 3.4:

 

Table 3.4 –Effect of agglomerated grains on Grinding performance

Wheel type	MRR gm/min	GR	Remarks
Standard	5.7	6.15
With organic agglomerates	7.1	4.02	Around 25% improvement in MRR over standard
With vitrified agglomerates	10.7	4.65	Around 50% improvement in MRR over standard

 

iii) Effect of  bond used for agglomeration on the grinding performance of belts are shown in Table 3.5:

 

Table 3.5 – Effect of bond used for agglomeration on the grinding performance of belts

Product type	Work removed (gm)	Metal removal rate (gm/min)	Wear of belt (gm)	Stretch ( mm)
AG36AA80org	387	33.0	20.9	10
AG36AA80vit	344	28.66	7.2	10

 

It is interesting to note that the bond used for agglomeration has a major impact on the performance of the final product. Comparing Table 3.4 and Table 3.5, it can be seen that vitrified agglomerates perform better in bonded applications whereas organic agglomerates perform better in coated products.

iv) Keeping track of developments - Patent mapping:

The user can keep track of such developments by continuous patent search method called patent mapping [128]. In this process, the user can search the website (www.uspto.gov.in) for the topic “abrasives” or “grinding”. Then based on the descriptions of the patent, one can classify the patent into a number of sub topics like a relationship matrix. Since there is provision for searching by patent numbers, one can continue the search from the next patent number from where he had last ended. This avoids duplication. A sample of patent mapping is shown in Table 3.6. A complete list covering about 150 patents on abrasives and grinding is listed in Appendix 1.

The table is developed on the following basis:
A patent search was done covering around 150 patents on abrasives.
The abstracts were scanned and classified under various headings.
The result was a subjective classification of the patents. This leads to the research trends in abrasive today.

Complete analysis is classified under different headings.

The headings used to classify are as below:
1. Grinding efficiency improvement
Grinding ratio
Metal removal rate
Design of grinding wheel

2. Grinding ratio
Improvements in mechanical property
Improvements in Grain/Bond holding property
External agents usage like grinding aids
3. Metal removal rate
Improving the mechanical property
Changing the grain friability
Using microcrystalline grain
Using composite grains
Only the headings with the classified patents are attached herewith.

Table 3.6 –Sample of patent mapping on grinding

 

Problem of invention			Solutions	Patents reference
Primary	Secondary	Tertiary
Grinding efficiency	Grinding Ratio	Wear	Improve the Mechanical Properties (Bending Strength, E-Modulus, Toughness) of Vitrified Bonds	Development of Glass-Ceramic Bonds	US 5573561, WO 9637342, EPO 636457, US 5711774, US 5407871, US 4526873, US 6123743, JP 2002018726, NZ 334347
				Development of Differential Fusibility	US 5536283, US  4898597, GB 572770, US 5863308, EPO 355630, GB 606441, US 5863308

 

3.1.4. Understanding from the study on grain types:

The study shows that even a small change in the chemistry of the grain (from regular brown to semi friable) can cause a major change in the abrasive performance.

The study on agglomerates shows that even in a given type of grain, the form in which the grain is present in the wheel has a major effect on its performance. If the grains are agglomerated, then the bond used for agglomeration tends to modify the fracture behaviour of the grain and hence the overall performance of the wheel is different from that of standard wheel of the same composition using unagglomerated grain.

Every type of grain is important and the ability to demand a particular type and form of the grain for a given application would be of advantage to the user. The user can keep track of such developments in abrasives by following a procedure called patent mapping.

3.2. Learning to assess safety of grinding wheels

3.2.1. Safety of grinding wheels:
Safety of grinding wheels is very important since speeds involved are very high of the order of 45 metres per second to as high as 100 metres per second.
Since vitrified wheels are liable to break if wrongly handled, it is important for the user to assess whether the wheels are free from any transit or handling cracks, before it is mounted on the machine.

As of now, all abrasive producers suggest only one method for the assessment of wheel safety- the ring test. In this test, the wheel is either supported on its edge or loosely held in the arbor with a finger and then tapped gently with an iron rod on its sides. If the wheel is good and free from crack, it will emit a metallic ring sound. If the wheel is defective, it will give a dull ring sound. The problem with this method is that it is highly subjective and there is no clear definition of metallic ring sound. Also, there are wheels like those made of organic bonds which do not respond to this ring test at all. As of now, there is a big gap in any useful and fool proof method for any user to assess his wheel condition as he receives it at his shopfloor.

3.2.2- Experimental plan:

Since the above test is very subjective and is bound to differ from person to person, an attempt was made to use the sound meter to measure the ring sound.
As part of ISO 14001, for measuring the sound level, for environmental safety purposes, almost every company has this instrument (sound meter). Hence the choice of this instrument. The details of the sound measuring instrument are in Table 3.7.
The purpose of this study is to understand the utility of the ordinary sound level meter for assessing the safety of a grinding wheel

Table 3.7 – Details of sound instrument

Sound instrument

Manufacturer: Northern Lab India

Type: Center 325 - Mini sound level meter, IEC651 , Type II Level : 3 ranges 32- 80 dB, 50 – 100dB, 80 – 130 dB

 

Three wheel gradings were chosen- two in alumina in grits 54 and 80 and the other in silicon carbide grit 400. This choice was just to cover the popular grit ranges and grain types.

Since the objective is to see whether defective wheels can be identified by sound values, it was necessary to generate defective wheels.  In each case, for generating defective wheels, peripheral cracks were introduced in the as molded green stage and then firing the wheels. Thus, in each of the three cases, we could have one good wheel (G) without cracks and one defective wheel (D), with cracks.

3.2.3. Measurement of sound index:

These wheels were characterised with sound meter. The meter had three ranges, namely 32-80, 50-100 and 80-130 decibels.
To have sufficient range, we used 80-130 decibels setting for all the tests.

Wheels are kept horizontal, supported on four flat strips formed like an X. These strips are in turn kept on a flat surface plate. The sound meter is just below the wheel, on the surface plate. The wheel is tapped gently with an ebonite hammer and the reading is noted.

The issue was that, immediately on switching on, the sound level meter was giving some reading - the ambient sound level. So, it was decided to take the difference of the sound value generated during tapping the wheel and the value displayed initially (ambient value before the test) as the sound index of the wheel. For example, if, on switching on, before the test, the reading is 43 decibels (dB) and if the reading observed on tapping the wheel is 84 dB, then the sound index of the wheel is taken as (84 minus 43) = 41 dB.

 As explained earlier, the sound reading before the test was noted and the wheel tapped gently with an ebonite hammer. The reading immediately after tapping was noted. The difference between the observed value and the initial value was taken as the sound index of the wheel. The values are shown in Table 3.8.

Table 3.8 - Sound index of 200 mm diameter wheel

 

Wheel type	Sound index in decibels
	Good wheels	Defective wheels
Aloxite grit 54	27	31
Aloxite grit 80	25.9	33.8
Silicon carbide grit 400	29.5	36.8

The above data shows that it is possible to use sound index effectively to identify good and bad wheels. Clearly, they have distinctly different values- good wheels always showing less than 30 decibels and cracked wheels always showing greater than 30 decibels.

Since vibration is dependent on dimensions of the wheel, this inference of 30 decibels may be considered only for wheels of diameter 200 mm and thickness 13 mm-used in these experiments.

It is expected that this cut off value may vary with wheel diameter and wheel thickness, possibly increasing when the wheel diameter increases. This is due to the fact that larger diameter wheels can vibrate well compared to smaller wheels.

3.2.4. Understanding from the study on sound index for safety:

By repeating this test on more varieties of wheels, and more number of wheels, it should be possible to arrive at one cut off value beyond which the wheel is unsafe and can be identified to be defective. For 200 mm diameter and 13 mm thick wheels, it appears that 30 dB could be the limit of safety, beyond which the wheel can be considered to be defective. Extensive work is required to establish this number for various sizes of wheels. However, for a given wheel being regularly used by any workshop, it is possible to establish a safe number by generating readings over a period of time, instead of using a subjective ring test and understanding the metallic sound.

3.3. Learning to characterise a grinding wheel

 

3.3.1. Characterisation of grinding wheels:

 

The word “Characterisation” is used to denote the ability to develop a unique number when wheel parameters are different. For example, in any wheel, it is the vitrified bond type that can be varied as refractory (hard) bond or fusible (soft) bond. Similarly, in a given bond type, the quantity of bond can be varied and is indicated as grade hardness denoted by the letters L, K, M  etc. As the alphabetical order increases, the wheel is said to be hard. Any of these changes will change the performance of the wheel. (These are described in Chapter 2).

Manufacturers of grinding wheels characterise the wheels using non-destructive methods like grindosonic, elastosonic etc. These tests operate by creating an impulse on the wheel and measuring the frequency of vibration. This frequency of vibration can be converted into young’s modulus and compared with that of the standard wheel. Also, only the supplier (manufacturer of the wheel) knows how to decipher the number obtained in such tests.

 

3.3.2. Experimental plan:

If any instrument is capable of uniquely identifying these changes in a wheel, then that instrument can be effectively used to indicate a possible performance of the grinding wheel. This is done by measuring the value for a wheel using the instrument, doing the actual grinding test with the wheel and then attempting to correlate the value obtained by the instrument to the grinding performance. Grindosonic and elastosonic are tests that are capable of differentiating such parameter changes.

In order to verify whether sound level meter also can be used similarly, wheels with three different hardness in one vitrified bond were made and characterised as stated above using sound instrument. To understand the range, two different ranges of the sound instrument were used to characterise the wheels. The wheels were subjected to grinding performance tests as described in Table 3.9

Table 3.9 - Test set up for grinding performance evaluation

Machine type	Cylindrical grinding
Work piece	EN19- 100 mm dia and 5 mm thick.
Wheel size	300 mm dia, 13 mm thick, 127 mm arbor
Test conditions	2mm/min – depth of cut
Grinding time	4 minutes test
Coolant	Cimcool

The wheels were evaluated for grinding ratio (GR) and metal removal rate (MRR). MRR indicates the volume of material removed per unit of time. GR refers to the volume of material removed per unit volume of the grinding wheel lost.

3.3.3. Observations and discussion:

 

The measured values are tabulated in Table 3.10.

 

Table 3.10 - Characterising by sound values for 300 mm  wheels

Wheel type	Frequency	Young’s modulus (Gpa)	GR	MRR gm/ min	Sound Index – dB
					32-80 range	80-130 range
RA 60 I5 VF8	2292	42.45	9.987	5.037	30.1	31
RA 60 K5 VF8	2116	49.56	14.436	5.265	30.4	35
RA 60 M5 VF8	2100	52.99	13.378	5.222	29.8	49

 

The above results show that sound index can be effectively used as a characterising method, just like elastosonic method, normally used by the manufacturers of grinding wheels. Sound index is able to clearly differentiate hard and soft wheels, just similar to elastosonic method. In both cases, as the wheel hardness increases, the observed value increases.

i) Impact of range used:
The low range of 32-80 dB gave very close results and could not effectively differentiate between grades. But the large range of 80-130 decibels was able to clearly differentiate between wheel grades as effectively as elastosonic method.

 

ii) Impact of wheel hardness:

Sound index is able to clearly differentiate hard and soft wheels, just similar to elastosonic method. In both cases, as the wheel hardness increases, the observed value increases. The above results show that sound index can be effectively used as a characterising method, just like elastosonic method.

iii) Ability to differentiate:

The benefit of any method is best understood by the amount of differentiation that it can bring, in differentiating the characteristic being measured. This can be analysed as in Table 3.10.

Observing the difference in units between the gradings, elastosonic shows a difference of 7 units between L and K  and around 3.4 units between K and L grades. This could be explained like this - as the wheel becomes harder, the vibration frequency drops and hence the Young’s modulus value, which is dependent on the natural frequency of vibration, also drops considerably.

In the chosen gradings, it appears that sound index is capable of showing larger variations between soft and harder gradings as we move harder. This could be because, harder wheels give stronger (more intense) sound when tapped, and hence register more decibels.
iv) Correlation to performance:

From the performance results, one can see that K5 grade has performed optimally for the chosen grinding conditions.

By continuous recording of sound index values of various wheels and registering their performance pattern, it is possible to arrive at a particular reference sound index of the wheel, which can provide the optimum grinding performance for the chosen set of standard conditions in any grinding shop. For example, in the chosen set of experiments done for this work, a sound index value of 48 and above would be required for getting a repeated standard performance range.

v) Reliability of results:

It should be borne in mind that the force of tapping, the distance of the sound level meter from the wheel and the influence of external sound are all variables that can affect the effectiveness of the present measurement.

• A free falling ebonite hammer from fixed position,
• fixtures to locate the instrument at fixed positions relative to the wheel, and
• a separate sound proof booth for testing,

will help in minimising the errors due to the above variables.

3.3.4. Understanding from the study on characterisation using sound index:
From the studies, it can be concluded that:

1. Sound measurement can be successfully used for characterising a grinding wheel.

2. With sufficient data of grinding, one can choose the wheels with a particular sound index for optimum performance.

3. Sound Index is capable of showing larger differentiation between harder and softer grades as compared to the grindosonic or elastosonic techniques, thereby making it a better tool for characterising.

4. Sound index equipment is far cheaper compared to grindosonic or elastosonic equipment and hence can be used economically.

3.4. Understanding equivalent gradings between different suppliers’ wheels

3.4.1. Need for equivalence:

Manufacturers of grinding wheels have different codes for indicating their grain types and combinations. There is no universally accepted code for wheel specification. While there is an accepted pattern that grain must come first grit must come next etc, there is no standard code for different type of grains. Therefore it becomes difficult for the user to understand the equivalent gradings when he wants to switch over from one supplier to the other.

One way is to browse the websites of various suppliers and try to understand the code used for various grain types, by reading the exhaustive description they give about their product and its performance. One such attempt was made scanning in detail, the websites of around 50 manufacturers and data on grain codification used by individual manufacturers is listed in Appendix 2.

 

3.4.2. Experimental plan:

 

It is also possible for the user to develop some test in a small test rig, get the sample for the small machine and evaluate the performance to assess whether the wheel is equivalent or not. In order to develop this method, wheels from various sources were obtained and performance test conducted as described in Table 3.9.

3.4.3. Observations and discussions:

 

The performance results are shown in Table 3.11

Table 3.11- Performance test for equivalence analysis

 

Grade	Grinding ratio	Metal removal rate- gm/min	Grinding efficiency	Specific Energy	Density (g/ci)
GNL-Mikron A60M6VCNM	15.828	5.795	15.135	1.096	37.6
Power A60N5V	17.002	5.594	15.863	1.072	36
Orient	14.254	5.651	21.712	0.656	36.3
CUMI A60	23.544	5.474	21.141	1.114	37.8

The procedure for calculations is given in Appendix 3.

It can be seen that density, which is one of the easily measurable parameter of the wheel is not really reflective of its performance. For example, wheels of GNL and that of CUMI are of almost same density but show different performance levels.

3.4.4. Understanding from the study on equivalence assessment of grinding wheels:

By such actual experiments, comparing the performance metrics like GR, MRR etc, the user can learn about equivalent wheels between different suppliers. This can be done over a period of time.

 

 

3.5. New learnings of this chapter

• The user must be aware of different grain types and grain forms in use for his kind of application. He can keep track of such developments by following the patent mapping procedure.

 

• Sound Index measurement can be effectively used to objectively assess the safety of the grinding wheel.

• Sound Index can be used to characterise and predict expected performance of the grinding wheel. If the user is familiar with this method, then this can be used for assessing equivalent wheels of different suppliers.

 

• When there is a need to switch between suppliers, the user must learn to assess equivalent gradings of grinding wheels. He can do so by
• referring to the Appendix 2  for grain codifications,
• using sound index as a reference point between different source wheels or
• actually comparing past performance of different source wheels using grinding performance metrics.