Materials and Design 53 (2014) 383–397

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier .com/locate /matdes

Technical Report

Reliability of mechanical properties of induction sintered iron basedpowder metal parts

0261-3069/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.matdes.2013.07.034

⇑ Corresponding author. Tel.: +90 2362012381; fax: +90 2362412143.E-mail addresses: can.civi@cbu.edu.tr (C. Çivi), necati.tahrali@cbu.edu.tr (N.

Tahralı), enver.atik@cbu.edu.tr (E. Atik).

Can Çivi ⇑, Necati Tahralı, Enver AtikCelal Bayar University, Engineering Faculty, Department of Mechanical Engineering, 45140 Muradiye-Manisa, Turkey

a r t i c l e i n f o

Article history:Received 21 March 2013Accepted 10 July 2013Available online 23 July 2013

Keywords:Powder metallurgySinteringInduction sinteringMechanical propertiesMicroVickers hardnessReliability

a b s t r a c t

Reliability and safety are important for machine and construction elements. In this study, iron basedpowder metal parts (3% Cu, 0.5% Graphite and 1% Kenolube lubricant by weight) were sintered at1200 �C by medium frequency induction sintering mechanism (30 KW powered and 30 kHz frequency).Mechanical property values of components were determined according to changing sintering time. Threepoint bending, % maximum strain, MicroVickers hardness (HV) and Rockwell-B hardness tests wereapplied. Statistical distribution functions were drawn and ultimate strength, ultimate strain, MicroVick-ers and Rockwell-B hardness values were determined depending on various reliability. As a result of theexperiments, it was concluded that, the hardness of powder metal materials should not be based onMicroVickers hardness.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Powder metal parts are widely used as machinery and construc-tion elements. Powder metallurgy is one of the highly preferredproduction method because of its widely advantages. Powderswhich have different composition are pressed and then sinteredat this method. Sintering is one of the most important issues ofpowder metallurgy because sintering causes significantly an in-crease in resistance of pressed powders [1]. The sintering processis generally performed in the sintering furnaces. It is done in batchfurnaces and continuous furnaces [1]. In addition, induction sinter-ing method is an important alternative of conventional sinteringmethod. The advantage of this process allows very quick densifica-tion to near theoretical density and prohibition of grain growth [2].Sintering and additional heat treatments of powder mixtures gen-erate the microstructure to meet the performance as required [3].

Sintered materials are generally have a porous structure, withthe increase in the amount of porosity, powder metal parts be-come brittle [1]. Mechanical Strength measurements of brittlematerials or of metals under conditions where they behave in abrittle manner show a high variability of results which requiresstatistical analysis [4]. While classical construction method isbased on safety, statistical construction method is based on reli-ability. Reliability values are range from 0 to 1, so F + R = 1 has arelation from Failure and Reliability [5]. Reliability is characteris-

tic of an item, expressed by the probability that the item will per-form is required function under given conditions for a stated timeinterval [6].

Normal distribution, also called Gaussian distribution, a proba-bility distribution is very important in many areas. Normal distri-bution has important applications in engineering and reliability[7]. Normal distribution is used to model of the physical, mechan-ical or chemical properties of a variety of systems. Gas moleculevelocity, clothing, sound, tensile strength of aluminum alloy steels,capacity variation of electrical condensers, electrical power con-sumption in a field, generator output voltage and electrical resis-tance are shown as examples application of the normaldistribution [8]. In another application, it is using analysis of thematerials produced and analysis of their ability to perform of theirfunctions [7]. Also, it is usually considered a normal distribution forUltimate stress and nominal stress values [9–13].

Stress–Life diagrams can be drawn for for 0.1%. . .99.9% reliabil-ity. It is possible to realize the life values for 0.1%. . .99.9% reliabilityaccording to the load intensity [14]. It is not proper to realize thelife analysis considering 50% reliability, for a machine elementwhich has a vital importance. It is also possible to determine lifevalues by using 90–99.9% reliability. However, these values be-come more important only in the case of competition of firms[15]. Similarly life values, ultimate stress values of materials canbe found in the range of 0.1%. . .99.9% reliability [8]. Normal distri-bution is defined in the range of (�1, +1). However, the reliabilitytheory is dealt with periods of life of object. Therefore, the distribu-tion of objects on the duration of life is assumed to be in the range(0, +1) [16].

Fig. 1. Induction sintering mechanism.

Fig. 2. Cupper coil.

Fig. 3. Laser pyrometer.

Fig. 4. Sintered samples.

Table 1Three point bending results of 8.4 min sintered samples.

Sample number Ultimate stress (N/mm2) Ultimate strain (%)

1 425.94 3.92 510.26 3.933 484.45 3.624 510.78 3.795 509.75 3.96 653.86 4.637 575.55 3.58 524.99 3.6

Table 2Three point bending results of 15 min sintered samples.

Sample number Ultimate stress (N/mm2) Ultimate strain (%)

1 562.64 3.372 426.07 2.933 525.41 4.054 603.51 5.355 613.22 4.296 533.06 4.197 533.33 4.28 464.72 3.49

Table 3Three point bending results of 30 min sintered samples.

Sample number Ultimate stress (N/mm2) Ultimate strain (%)

1 467.26 3.242 539.43 3.793 579.15 4.44 517.78 4.285 577.69 5.046 653.92 5.577 570.05 4.668 534.15 4.81

384 C. Çivi et al. / Materials and Design 53 (2014) 383–397

In previous studies, induction sintering was generally carriedout with high frequency induction sintering unit at the same timepressing process (HFIFS) [17–23]. In this study, sintering was

carried out after pressing process with medium frequency induc-tion unit (30 kHz). In general, the average values of experimentalresults were given at powder metallurgy studies. Instead of aver-age values of mechanical properties, most of the time mechanicalproperties values which have high reliability are more importantat usage areas of powder metal parts. In this study, mechanicalproperties of powder metal parts were determined with mechani-cal tests. The results of these tests were evaluated statistically andthe results which have 10%, 50% and 90% reliability were identifiedand compared with each other. Reliability analysis of ultimatestress, ultimate strain, Rockwell-B hardness and Vickers hardnesswere done. The most suitable sintering time according to the

Table 4Comparison of three-point bend test results of the samples sintered in three different times.

Time ½ru�R10 ðN=mm2Þ ½ru�R50 ðN=mm2Þ ½ru�R90 ðN=mm2Þ ½eu�R10 ð%Þ ½eu�R50 ð%Þ ½eu�R90 ð%Þ

8.4 min 609.985 524.44 438.8949 4.3074 3.8587 3.409915 min 618.1673 532.74 447.312 4.9248 3.9837 3.042530 min 624.8154 554.92 485.0245 5.4046 4.4737 3.5427

C. Çivi et al. / Materials and Design 53 (2014) 383–397 385

statistical evaluation was obtained with reliability. Also, as a resultof Rockwell-B and MicroVickers hardness measurements, it wasdetermined that Micro-Vickers hardness measurement is notappropriate for sintered components due to the small dimensionsof the MicroVickers diamond pyramid hardness tester and porosityand alloying elements of powder metal parts.

2. Experimental studies

In this study, Högenas ASC 100.29 iron powder (3% Cu, 0.5%Graphite and 1% kenolube lubricant by weight, size of 45–106 lm.) was used. Metal powder was pressed under 600 N/mm2

with one axis press and 10 � 10 � 55 mm about 37 gr samples

0 200 400 600

0.002

0.004

0.006

0.008

0 200 400 600

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 5. Normal and cumulative distribution curves of u

were formed. Samples were sintered in atmosphere with mediumfrequency induction sintering unit for 8.4, 15 and 30 min at1120 �C to compare (Fig. 1). Induction sintering was carried outin heat resistant glass in cupper coil (Fig. 2).Sintering temperaturewas kept constant at 1120� with laser pyrometer (Fig. 3). Sinteredsamples were given Fig. 4.

Ultimate stress and ultimate strain values of sintered samplesare determined by three point bending test. Instead of6.35 � 12.7 � 31.7 mm transverse rapture strength sample accord-ing to ASTM: B 528-12, 10 � 10 � 55 mm samples produced ASTME23-12c for more homogenous distribution of flow line and heat-ing of the samples. 3 Point bending test was performed on the sam-ples by using Autograph Shimadzu AG-IS 100 kN universal test

800

800

ltimate stress values for 8.4 min sintered samples.

0 2 4 6 8

0.2

0.4

0.6

0.8

1.0

1.2

0 2 4 6 8

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 6. Normal and cumulative distribution curves of ultimate strain values for 8.4 min sintered samples.

386 C. Çivi et al. / Materials and Design 53 (2014) 383–397

machine. Rockwell-B hardness and MicroVickers hardness test wasapplied to samples. MicroVickers hardness values were measuredwith Future-Tech FM-700 microhardness test machine by loading50 gr of force for 10 s according ASTM: E384-11e1. The results ofthe experiments were evaluated for each test result, distributionfunctions were drawn and according to the distribution function;ultimate stress, ultimate strain and hardness values were deter-mined. Rockwell-B test was applied according to ASTM E18-12.

Distribution functions were drawn based on the fallowing for-mulas with computational software program,

FðxÞ ¼Z

1s�

ffiffiffiffiffiffiffi2pp Exp � x� �xð Þ

2�s2x

� �ð1Þ

FðxÞ ¼Z

1s�

ffiffiffiffiffiffiffi2pp Exp �ðx�

�xÞ2�s2

x

� �ð2Þ

X: Mechanical property value.Sx: Standard deviation of mechanical property value.Maximum and minimum values of ultimate stress (ru), ulti-mate strain (eu), MicroVickers (HV), Rockwell-B (HRB) deter-mined with following formulas. Z values were taken from Ztable [24].

½xMax� ¼ �xþ z�sx ð3Þ

½xmin� ¼ �x� z�sx ð4Þ

3. Results and discussion

3.1. Three point bending test results

The results of three point bending experiments of the 8 samplesfor each sintering time can be seen in Tables 1–3, and comparisonof three-point bending test results of the samples sintered in threedifferent times according to 10% and 90% reliability values can beseen Table 4.

Samples sintered for 8.4 minru ¼ 524;44 N=mm2 (The average of ultimate stress values,

[ru]R50)

eu ¼ 3:8587 ð%Þ (The average of ultimate strain values, [eu]R50).sr = 66.8321 (The standard deviation of ultimate stress values).z = 1.28 from z table, for 90% (R90) and 10% reliability values(R10).

0 200 400 600 800

0.002

0.004

0.006

0.008

0 200 400 600 800

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 7. Normal and cumulative distribution curves of ultimate stress values for 15 min sintered samples.

C. Çivi et al. / Materials and Design 53 (2014) 383–397 387

½ru�R90 ¼ ru � z�sr ¼ 524:44� 1:28 � 66:8321

¼ 438:8949 N=mm2 ð5Þ

½ru�R10 ¼ ru þ z�sr ¼ 524:44þ 1:28 � 66:8321

¼ 609:985 N=mm2 ð6Þ

se = 0.3506 (the standard deviation of ultimate strain values).

½eu�R90 ¼ eu � z�se ¼ 3:8587� 1:28�0:3506 ¼ 3:4099 ð7Þ

½eu�R10 ¼ eu þ z�se ¼ 3:8587þ 1:28�0:3506 ¼ 4:3074 ð8Þ

Samples sintered for 15 min

ru ¼ 532:74 N=mm2

eu ¼ 3:9837%

sr ¼ 66:7401

½ru�R90 ¼ rk � z�sr ¼ 532:74� 1:28�66:7401

¼ 447:312 N=mm2 ð9Þ

½ru�R10 ¼ rk þ z�sr ¼ 532:74þ 1:28�66:7401

¼ 618:1673 N=mm2 ð10Þ

se ¼ 0:7353

½eu�R90 ¼ e� z�se ¼ 3:9837� 1:28�0:7353 ¼ 3:0425 ð11Þ

½eu�R10 ¼ �eþ z�se ¼ 3:9837þ 1:28�0:7353 ¼ 4:9248 ð12Þ

Samples sintered for 30 min

ru ¼ 554:9281 N=mm2

eu ¼ 4:4737%

sr ¼ 54:6058

½ru�R90 ¼ rk � z�sr ¼ 554:92� 1:28�54:6058

¼ 485:0245 N=mm2 ð13Þ

0 2 4 6 8

0.2

0.4

0.6

0.8

1.0

1.2

0 2 4 6 8

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 8. Normal and cumulative distribution curves of ultimate strain values for 15 min sintered samples.

388 C. Çivi et al. / Materials and Design 53 (2014) 383–397

½ru�R10 ¼ rk þ z�sr ¼ 554:92þ 1:28�54:6058

¼ 624:8154 N=mm2 ð14Þ

se ¼ 0:7273

½eu�R90 ¼ e� z�se ¼ 4:4737� 1:28�0:7273 ¼ 3:5427 ð15Þ

½eu�R10 ¼ �eþ z�se ¼ 4:4737þ 1:28�0:7273 ¼ 5:4046 ð16Þ

It is shown of the tables, by increase of sintering time, the ulti-mate stress and ultimate strain values with 10%, 50% and 90% reli-ability percent are increased.

3.2. Normal distribution curves of three point bending test results

Ultimate stress distribution function curves of samples sinteredfor 8.4 min were given Fig. 5

�x ¼ rk; �x ¼ 524:44

Sx ¼ sr; Sx ¼ 66:83

Ultimate strain distribution function curves of samples sinteredfor 8.4 min were given Fig. 6

�x ¼ e; �x ¼ 3;85Sx ¼ Se; Sx ¼ 0;3506

Ultimate stress distribution function curves of samples sintered for15 min were given Fig. 7

�x ¼ rk; �x ¼ 532;74Sx ¼ sr; Sx ¼ 63:74

Ultimate strain distribution function curves of samples sinteredfor 15 min were given Fig. 8

�x ¼ e; �x ¼ 3:98Sx ¼ Se; Sx ¼ 0:7353

Ultimate stress distribution function curves of samples sinteredfor 30 min were given Fig. 9

�x ¼ rk; �x ¼ 534:14Sx ¼ sr; Sx ¼ 54:60

Ultimate stress distribution function curves of samples sinteredfor 30 min were given Fig. 10

�x ¼ e; �x ¼ 4:81Sx ¼ Se; Sx ¼ 0:7253

0 200 400 600 800

0.002

0.004

0.006

0.008

0 200 400 600 800

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 9. Normal and cumulative distribution curves of ultimate stress values for 30 min sintered samples.

C. Çivi et al. / Materials and Design 53 (2014) 383–397 389

3.3. MicroVickers hardness measurement results

Samples sintered for 8.4 minHV ¼ 245:4854 (the average of Micro-Vickers hardness values,

½HV �R50).sHV = 28.1622 (the standard deviation of MicroVickers hardness

values).

½HV �R90 ¼ HV � z�sHV ¼ 245:4854� 1:28 � 28:1622

¼ 209:4377 ð19Þ

½HV �R10 ¼ HV þ z�sHV ¼ 245:4854þ 1:28 � 28:1622

¼ 281:5330 ð20Þ

Samples sintered for 15 min

HV ¼X8

i¼1

HV ¼ 258:8104

sHV ¼ 28:3241

½HV �R90 ¼ HV � z�sHV ¼ 258:8104� 1:28 � 24:3241

¼ 227:6755 ð21Þ

½HV �R10 ¼ HV þ z�sHV ¼ 258:8104þ 1:28 � 24:3241

¼ 289:9453 ð22Þ

Samples sintered for 30 min

HV ¼X8

i¼1

HV ¼ 244:3417

sHV ¼ 25:5475

½HV �R90 ¼ HV � z�sHV ¼ 244:3417� 1:28 � 25:5475

¼ 211:6409 ð23Þ

½HV �R10 ¼ HV þ z�sHV ¼ 244:3417þ 1:28 � 25:5475

¼ 277:0425 ð24Þ

0 2 4 6 8

0.2

0.4

0.6

0.8

1.0

1.2

0 2 4 6 8

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 10. Normal and cumulative distribution curves of ultimate strain values for 30 min sintered samples.

390 C. Çivi et al. / Materials and Design 53 (2014) 383–397

3.4. Normal distribution curves of MicroVickers hardness test results

MicroVickers distribution function curves of samples sinteredfor 8.4 min were given Fig. 11

�x ¼ HV ; �x ¼ 245:48Sx ¼ SHV ; SHV ¼ 28:1622

MicroVickers distribution function curves of samples sinteredfor 15 min were given Fig. 12

�x ¼ HV ; �x ¼ 258:81Sx ¼ SHV ; SHV ¼ 24:3241

MicroVickers distribution function curves of samples sinteredfor 30 min were given Fig. 13 (see Tables 5–8)

�x ¼ HV ; �x ¼ 244:34Sx ¼ SHV ; SHV ¼ 25:5475

3.5. Rockwell-B hardness measurement results

Rockwell-B hardness test results of samples are shown Tables9–12

Samples sintered for 8.4 min

HRB ¼ 54:3625 HRB (the average of Rockwell-B Hardness val-ues, [HRB]R50).sHRB = 2.5011 (the standard deviation of Rockwell-B Hardnessvalues).

½HRB�R90 ¼ HRB� z�sHV ¼ 54:3625� 1:28�2:5011 ¼ 51:161 ð25Þ

½HRB�R10 ¼ HRBþ z�sHV ¼ 54:3625þ 1:28�2:5011

¼ 57:5639 ð26Þ

Samples sintered for 15 min

HRB ¼ 55:5937HRB

sHRB ¼ 2:8609

0 20 40 60 80 100

0.05

0.10

0.15

0.20

0 20 40 60 80 100

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 11. Normal and cumulative distribution curves of MicroVickers values for 8,4 min sintered samples.

C. Çivi et al. / Materials and Design 53 (2014) 383–397 391

½HRB�R90 ¼ HRB� z�sHV ¼ 55:5937� 1:28�2:8609

¼ 51:9317 ð27Þ

½HRB�R10 ¼ HRBþ z�sHV ¼ 55:5937þ 1:28�2:8609

¼ 59:2556 ð28Þ

Samples sintered for 30 min

HRB ¼ 59:6125HRB

sHRB ¼ 5:3954

½HRB�R90 ¼ HRB� z�sHV ¼ 59:6125� 1:28�5:3954

¼ 52:7063 ð29Þ

½HRB�R10 ¼ HRBþ z�sHV ¼ 59:6125þ 1:28�5:3954

¼ 66:5186 ð30Þ

3.6. Normal distribution curves of Rockwell-B hardness test results

Rockwell-B distribution function curves of samples sintered for8.4 min were given Fig. 14

�x ¼ H; H ¼ 54:36Sx ¼ SH; SH ¼ 2:5011

Rockwell-B distribution function curves of samples sintered for15 min were given Fig. 15

�x ¼ H; H ¼ 55:59Sx ¼ SHRB; SHRB ¼ 2:8609

Rockwell-B distribution function curves of samples sintered for30 min were given Fig. 16

�x ¼ H; ¼59:13Sx ¼ SH; SH ¼ 6:0197

0 20 40 60 80 100

0.05

0.10

0.15

0.20

0 20 40 60 80 100

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 12. Normal and cumulative distribution curves of MicroVickers values for 15 min sintered samples.

392 C. Çivi et al. / Materials and Design 53 (2014) 383–397

Khalil and Almajid [25] reported that the compressive strengthwas significantly improved with increasing sintering time up to3 min and then decreased after 4 min of sintering with high fre-quency induction sintering of nanostructured magnesium/hydroxyapatite nanocomposites. In this study, the increasing ofstrength of materials was continued up to 30 min unlike Khaliland Almajid [25]. Çavdar and Atik [26] reported that strengthand Microhardness values of iron based powder metal materialsin the medium frequency induction sintering process (30–50 kHz) were increased with sintering time. Mechanical tests re-sults were decreased via decreasing temperature 900–1200� andsintering time up to 700 s (11.66 min). In this study, induction sin-tering time was increased up to 30 min (the same time as conven-tional sintering time in furnace) and it was seen that mechanicalproperty values were increased with sintering time. Çivi and Atik[27] reported that Maximum stress of the 8.4 min medium fre-quency induction sintered Fe based samples were caught andpassed samples sintered 30 min by classic sintering furnace and

by increase of sintering time, maximum stress and maximumstrain values of samples were increased. Zhang and Sandström[28] investigated the effects of sintering temperature, time andatmosphere on the properties of classically sintered steels withthese Fe–Mn–Si master alloy powders. Eventually, they found thedensity of the compacts increased with sintering temperatureand time. The ultimate tensile strength and hardness increasedwith sintering temperature and time mainly due to increasingamounts of bainite and martensite after cooling. Elongation wasinitially raised with sintering temperature and time probably dueto improved bonding between powder particles. And also theysaw liquid phase sintering accelerates the sintering process, whichleads to improved mechanical properties. In this study, mediumfrequency induction sintering was done. As a result of this study,mechanical property values of induction sintered powder metalsamples except Vickers hardness results were generally increasedwith sintering time up to 30 min classically sintering time in fur-nace likely Khalil and Almajid [25], Çavdar and Atik [26] and Çivi

0 20 40 60 80 100

0.05

0.10

0.15

0.20

0 20 40 60 80 100

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 13. Normal and cumulative distribution curves of MicroVickers values for 30 min sintered samples.

Table 5Vickers microhardness test results of 8.4 min sintered samples.

Sample number 1 2 3 4 5 6 7 8

Average (HV) 242.97 196.13 236.13 268.83 282.93 228.18 271.93 236.77

Table 6Vickers microhardness test results of 15 min sintered samples.

Sample number 1 2 3 4 5 6 7 8

Average (HV) 260.22 236.10 265.60 221.07 247.93 283.77 296.38 259.42

Table 7Vickers microhardness test results of 30 min sintered samples.

Sample number 1 2 3 4 5 6 7 8

Average (HV) 210.25 273.05 230.27 214.92 261.28 231.23 262.13 271.60

Table 8Comparison of Vickers microhardness test results of the samples sintered in three different durations.

Time [HV]R10 [HV]R50 [HV]R90

8.4 min 281.5331 245.4854 209.437815 min 289.9453 258.8104 227.675530 min 277.0426 244.3417 211.6408

C. Çivi et al. / Materials and Design 53 (2014) 383–397 393

Table 9Rockwell-B hardness test results of 8.4 min sintered samples.

Sample Number 1 2 3 4 5 6 7 8

Average (HRB) 51.9 55.875 52.75 55.325 54.125 51.55 59.25 54.125

Table 10Rockwell-B hardness test results of 15 min sintered samples.

Sample number 1 2 3 4 5 6 7 8

Average (HRB) 59 54.65 53.875 56.15 59.25 51.025 57.3 53.5

Table 11Rockwell-B hardness test results of 30 min sintered samples.

Sample number 1 2 3 4 5 6 7 8

Average (HRB) 65.25 57.675 60.575 55.375 55 67.875 50.425 62.65

Table 12Comparison of Rockwell-B test results of the samples sintered in three different durations.

[HRB]R10 [HRB]R50 [HRB]R90

8.4 min 57.5639 54.3625 51.16115 min 59.2556 55.59375 51.931730 min 66.5186 59.6125 52.7063

0 100 200 300 400 500

0.002

0.004

0.006

0.008

0.010

0.012

0.014

100 200 300 400 500

0.5

0.5

1.0

Fig. 14. Normal and cumulative distribution curves of Rockwell-B values for 8.4 min sintered samples.

394 C. Çivi et al. / Materials and Design 53 (2014) 383–397

0 100 200 300 400 500

0.005

0.010

0.015

100 200 300 400 500

0.5

0.5

1.0

Fig. 15. Normal and cumulative distribution curves of Rockwell-B values for 15 min sintered samples.

C. Çivi et al. / Materials and Design 53 (2014) 383–397 395

and Atik [27]. The increasing of mechanical property values werecontinued up to 30 min maximum sintering time contrarily Khaliland Almajid [25].

Many powder metal and induction sintering studies,MicroVickers hardness test have been using. In this study, it issuggested from results of the statistical investigation of teststhat MicroVickers hardness is not appropriate for the powdermetal parts which have porosity and alloying element. Also, inthis study, statistical investigation of experimental results wasperformed unlike other studies and values which have differentreliability were obtained. Thus, the mechanical property valuesthat investigated in this study, which have different reliabilitydepending on usage areas of powder metal parts could be usedin industry.

4. Conclusions

Generally, average mechanical property values of powdermetal parts are used in the studies and catalogs. In this study,

statistical investigations of mechanical properties of inductionsintered powder metal parts were done. The mechanical proper-ties were investigated in normal distribution. Mechanical prop-erty values can be determined in the range of 0.1%. . .99.9%reliability by severity of usage area of powder metal component.As a result of,

� Mechanical properties of the samples in any reliability are pos-sible to read from graphics of normal distribution and cumula-tive distribution functions. In this way, a reliability value whichis determined the area of usage of parts can be seen from thegraphics.� The mechanical properties with 90% reliability have great

importance in the use of powder metal parts because of safety.These values can be read of the graphics. Also the graphics pro-vide to determine corresponding to 0–99.9% reliability and fail-ure values.� The percentage of damaged parts at a certain value of stress,

strain and hardness can be determinable with graphics ofcumulative statistical distribution functions.

0 100 200 300 400 500

0.005

0.010

0.015

100 200 300 400 500

0.5

0.5

1.0

Fig. 16. Normal and cumulative distribution curves of Rockwell-B values for 30 min sintered samples.

396 C. Çivi et al. / Materials and Design 53 (2014) 383–397

� By increase of sintering time, the ultimate stress, ultimatestrain, Rockwell-B hardness values with 10%, 50% and 90% reli-ability percent were increased at this study.� It was also found that Vickers hardness values were not

increased in generally with induction sintering time. It couldbe obtained because of porosities, alloying elements and smalldimensions of the MicroVickers diamond pyramid hardness tes-ter. It is suggested that MicroVickers hardness test is not appro-priate for the powder metal parts which have porosity andalloying element. Due to results of the Rockwell-B hardnesstests, it is also suggested that macrohardness tests such asRockwell-B and Brinell are more appropriate for powder metalparts which have porosities and alloying elements because ofbigger dimensions of these hardness testers.

References

[1] Randall MG. In: Durlu N, Sarıtas� S, Türker M, editors. Powder metallurgy andparticulate materials processing. Ankara/Turkey: TMMD; 2007.

[2] Shon IJ, Jeong IK, Ko IY, Doh JM, Woo KDo. Sintering behavior and mechanicalproperties of WC–10Co, WC–10Ni and WC–10Fe hard materials produced byhigh-frequency induction heated sintering. Elsevier Sci, Ceram Int2009;35:339–44.

[3] Narasimhan KS. Sintering of powder mixtures and the growth of ferrouspowder metallurgy. Mater Chem Phys 2001;67:56–65.

[4] Dieter G. Mechanical Metallurgy. 3rd ed. McGraw-Hill Book Company Science/Engineering/Math; 1986.

[5] Tahrali N, Dikmen F. Reliability and life calculations for machineelements. Yıldız Technical University Press; 2004.

[6] Ebelign CE. Reliability engineering: theory and practice. 6thed. Zürich: Springer; 2010.

[7] Türkan AH. Statistical distribution models used in reliability analysis. MSc.Thesis, Department of Statistics, Institute of Natural and Applied Sciences,University of Çukurova; 2007.

[8] Ireson WG, Coombs CF, MOSS RY. Handbook of reliability engineering andmanagement. McGraw-Hill; 1996.

[9] Akkurt M. Principles of reliability at machine construction. _Istanbul: BirsenPress; 1997.

[10] Lipson C, Sheth NJ. Engineering experiments. New York: McGraw-Hill BookCompany; 1973.

[11] Haugen BE, Wirsching PH. Probabilistic design. J Mach Des; Number9,11,12,13,14; 1975.

[12] Wirsching PH, Kempert IE. MD series on fatigue; a fresh look at fatigue; modelsthat predict fatigue failure; fatigue failure in the real world. J Mach Des,Number: 12,16,19; 1976.

[13] Pronikov AS. Dependability and durability of engineeringproducts. London: Butterworths; 1973.

[14] Bayraktar M, Tahralı N, Güçlü, R. Reliability and fatigue life evaluation ofrailway axles. J Mech Sci Technol 2010;24(3):671–9.

[15] Dikmen F, Bayraktar M, Güçlü R. Railway axle analyses: fatigue damage andlife analysis of rail vehicle axle. J Mech Eng 2012;58:545–52.

[16] Gnedenko B, Ushakov I. Probabilistic reliability engineering. In: James Falk. AWiley-Interscience, Publication: John Wiley & Sons, Inc. New York, Chichester,Brisbane,Toronto, Singapore; 1995.

[17] Kim CH, Yoon JK, Doh JM, Ko IY, Shon IJ. Rapid sintering process andmechanical properties of binderless ultra fine tungsten carbide. Mater Sci EngA 2006:435–6. 717–724.

C. Çivi et al. / Materials and Design 53 (2014) 383–397 397

[18] Shon IJ, Jeong IK, Jeong HV, Park JH, Kim BR, Lee KT. Effect of Fe2O3 addition onconsolidation and properties of 8 mol% yttria-stabilized zirconia by high-frequency induction heated sintering (HFIHS). Ceram Int 2009;35:363–8.

[19] Khalil AK, Almajid AA. Effect of high-frequency induction heat sinteringconditions on the microstructure and mechanical properties of nanostructuredmagnesium/hydroxyapatite nanocomposites. Mater Des 2012;36:58–68.

[20] Kim HC, Oh DY, Shon IJ. Sintering of nanophase WC-15 vol. %Co hard metals byrapid sintering process. Refractorry Met Hard Mater 2004;22:197–203.

[21] Kim HC, Shon IJ. Rapid sintering of ultra-fine WC-10 wt% Co by high frequencyinduction heating. J Mater Sci 2005;40:2849–54.

[22] Khalil KA, Kim SW. Mechanical wet-milling and subsequent consolidation ofultra-fine Al2O3-(ZrO2+Y2O3) bioceramics by using high-frequency inductionheat sintering. Science Press. Trans Nonferrous Met Soc China 2007;17:21–6.

[23] Park HK, Shon IJ, Yoon JK, Doh JM, Ko IY, Munir ZA. Simutaneous synthesis andconsolidation of nanostructured NbSi2-Si3N4 composite from mechanically

activated powders by high frequency induction-heated combustion. J AlloysComp 2008;461:560–4.

[24] Montgomery DC, Runger GC. Applied statistics and probability for engineers.3rd ed. United States of America: John Wiley & Sons, Inc.; 2002.

[25] Khalil KA, Almajid AA. Effect of high-frequency induction heat sinteringconditions on the microstructure and mechanical properties of nanostructuredmagnesium/hydroxyapatite nanocomposites. Mater Des 2012;36:58–68.

[26] Çavdar U, Atik E. Determination sintering time and temperature parametersfor the medium frequency induction sintering of iron based powder metalcompacts. Wulfenia J 2013;20:273–83.

[27] Çivi C, Atik E. Comparison of effect of induction and classical sintering tomechanical properties of powder metal components. In: AIP conferenceproceedings, vol. 1476; 2012. p. 119–22.

[28] Zhang Z, Sandström R. Fe–Mn–Si master alloy steel by powder metallurgyprocessing. J Alloys Comp 2004;363:194–202.