Journal of Materials Engineering and Performance

, Volume 16, Issue 2, pp 242–247

Warpage Behavior of 7075 Aluminum Alloy Extrusions

Authors

    • Mechanical Engineering DepartmentLoyola Marymount University
  • T.M. Ruperto
    • Mechanical Engineering DepartmentLoyola Marymount University
  • S.L. Vasquez
    • Mechanical Engineering DepartmentLoyola Marymount University
  • A.Y. Yue
    • Mechanical Engineering DepartmentLoyola Marymount University
  • D.J. Manriquez
    • Mechanical Engineering DepartmentLoyola Marymount University
  • J.C. Quilla
    • Mechanical Engineering DepartmentLoyola Marymount University
  • S.H. Harris
    • Mechanical Engineering DepartmentLoyola Marymount University
  • S. Hannan
    • Mechanical Engineering DepartmentLoyola Marymount University
  • J. Foyos
    • Mechanical Engineering DepartmentLoyola Marymount University
  • E.W. Lee
    • Naval Air Systems Command
  • B. Pregger
    • Naval Air Systems Command
  • N. Abourialy
    • Naval Air Systems Command
  • J. Ogren
    • Mechanical Engineering DepartmentLoyola Marymount University
Article

DOI: 10.1007/s11665-007-9044-0

Cite this article as:
Es-Said, O., Ruperto, T., Vasquez, S. et al. J. of Materi Eng and Perform (2007) 16: 242. doi:10.1007/s11665-007-9044-0

Abstract

Extruded I sections of 7075-T6 aluminum were machined into four different sections shapes: L, short depth L, T, and short depth T. The furnace was preheated to 416 °C (780 °F) and the samples were placed inside. The temperature was raised to 471 °C (880 °F) and then the samples were quenched in either a 30% polyalkylene Glycol solution or water, both at 15 °C (59 °F). Points on the distorted samples were recorded before and after the solution treatment; the difference between the measurements indicated the extent of warpage.

Keywords

7075 extrusionsvarying section shapeswarpage

Introduction

Heat treatable aluminum alloys are quenched from the solution treating temperature, typically between 466 °C (870 °F) and 566 °C (1050 °F), to achieve a supersaturated solid solution prior to controlled artificial age hardening, (Ref 1, 2). Agitated cold water quenching often results in distortion (warpage) of the parts and development of high residual stresses, (Ref 1-3). The high cooling rate produces large temperature gradients between thick and thin sections which causes localized plastic flow, which in turn results in warpage after quenching or during machining, (Ref 1).

Warpage is a major problem, which leads to laborious, expensive, and time consuming straightening operations and in some cases to scrapping of large expensive parts, (Ref 4). Warpage can be minimized by adding polymers to water quenchants to reduce the convective or film coefficient between the part and the water, (Ref 5-8). These synthetic quenchants retard the heat transfer from the components surface and reduce the temperature differential between different thicknesses of the material, (Ref 1, 2, 4).

Much work has been reported in efforts to reduce warpage, (Ref 6-9), however, systematic studies are needed to characterize warpage profiles as a function of manufacturing method (rolled plates, extrusions, forgings...), variations in thickness of similar shapes, variation in thickness of a component, section shape, and the like.

Foyos et al. (Ref 10) determined the warpage of 7050 aluminum alloy I and C sections by a height gage and surface plate, Konyukov et al. (Ref 11) measured the warpage of aluminum samples from the deviation from the plane of the supporting plate at different points along the perimeter and Maidment (Ref 12) measured the flatness of the samples with a Zeiss SMM instrument at several points.

In this paper an experiment is carried out to determine the effect of section shape and web depth on the warpage behavior of 7075 extrusions. The effects of two quenchants: water (high film coefficient) and 30% polyalkylene glycol solution (low film coefficient) on the warpage behavior are also studied.

Experimental Procedure

I section extrusions of 7075-T6 aluminum alloy of 0.335 mm (0.0132 in) flange thickness, 5.6 cm (2.2 in) flange width, 0.28 cm (0.11 in) wall thickness, 17.8 cm (7.0 in) depth of the web and 30.5 cm (12 in) length were machined into L and T sections and short depth 8.9 cm (3.5 in) L and T sections. The quantitative profiles of 30 samples of five shapes (I, L, T, short depth L and T) were characterized by a Brown and Sharpe coordinate measuring machine. In this method, a point at one corner of the beam was used as a reference, labeled zero (Ref 10). The beams were clamped at the reference point and held fixed on a flat and the height was measured along the Z-axis at several points.

The extrusions of different shapes, Fig. 1(a) were placed in a furnace preheated to 416 °C (780°F), and the temperature was raised by 27.5 °C (50 °F) per hour to 471 °C for two more hours. The solution treatment proceeded at 471 °C (880°F) for two more hours. A set of three samples per shape, (I, L, T, short L and short T) was quenched in 30% glycol solution surrounded by ice at 15 °C (59 °F) and another set was quenched in water at the same temperature. Quenching was along the length of the extrusion at a rate of 7.6 cm/s (3′′/s). The temperature of the quenchant and the quenching rate were chosen to amplify the warpage effect similar to the procedure of an earlier work, (Ref 10). The quantitative profiles of the samples were again characterized by the Brown & Sharpe coordinate machine. The difference between the profile measurements at different points and the angle between the web and flange in each sample before and after solution treatment quantified the amount of warpage.
https://static-content.springer.com/image/art%3A10.1007%2Fs11665-007-9044-0/MediaObjects/11665_2007_9044_Fig1_HTML.gif
Fig. 1

(a) Extrusions of different shapes as machined. (b): Extrusions after solution treatment and quenching

Results and Discussion

Section Shapes

The warpage profiles for the I, L and T sections are also shown schematically in Fig. 1(b).

I Sections

For the I section, Fig. 2(a), points 1-3, and 10-12 are located on the inner part of the bottom flange, points 4-9 on the outer part of the top flange and points 13-21 on the side of the web. Points 1 and 3 are 2.54 cm (1 in) away from the edges and point 2 is 15.2 cm (6 in) from both edges. Points 4-6, 7-9 and 10-12 are distributed in a similar pattern. Points 13, 16 and 19 are at the top, middle and bottom of the web, points 14, 17, 20, and 15, 18, 21 are distributed in a similar pattern. The Beams were placed in two configurations. In Fig. 2(a) the heights of all the points except 13-21 were measured along the depth of the web. In Fig. 2(b), point 13 is considered the zero position for the coordinate measuring machine. The probe is then moved to points 14-21 to identify the distortion in the length-depth of the web plane. The measurement of the same marked points was repeated after the solution treatment.
https://static-content.springer.com/image/art%3A10.1007%2Fs11665-007-9044-0/MediaObjects/11665_2007_9044_Fig2_HTML.gif
Fig. 2

I section profile in two configurations (a) and (b)

The distortion for the I sections was measured for the three samples quenched in water and three samples quenched in glycol. In Table 1, the measurements of two samples, one quenched in water and the other in Polyalkylene glycol are shown. The difference between the pre-measure and the remeasure values (as received and as solution treated) are smaller for glycol quenched samples. If the units are converted to inches, also included in Table 1, the difference is in tenths to hundredths of an inch for water quenched samples and mostly in hundredths of an inch for glycol quenched samples.
Table 1

The distortion in cms and in inches for water and glycol quenched I sections

Quenchant

Point

Pre-measure, in.

Pre-measure, cm

Remeasure, in.

Remeasure, cm

Difference, in.

Difference, cm

Water at 59 °F Sample 1

 

1

0.13111

0.3330194

0.14620

0.37135

0.01509

0.03833

2

0.13060

0.331724

0.16197

0.41140

0.03137

0.07968

3

0.13040

0.331216

0.19591

0.49761

0.06551

0.16640

4

7.19864

18.2845456

7.29414

18.52712

0.09550

0.24257

5

7.19922

18.2860188

7.30691

18.55955

0.10769

0.27353

6

7.20058

18.2894732

7.32947

18.61685

0.12889

0.32738

7

7.18216

18.2426864

6.93104

17.60484

−0.25112

−0.63784

8

7.18222

18.2428388

6.93703

17.62006

−0.24519

−0.62278

9

7.18054

18.2385716

6.98639

17.74543

−0.19415

−0.49314

10

0.13345

0.338963

0.13926

0.35372

0.00581

0.01476

11

0.13290

0.337566

0.14638

0.37181

0.01348

0.03424

12

0.13096

0.3326384

0.14398

0.36571

0.01302

0.03307

13

0.00451

0.0114554

0.20122

0.51110

0.19671

0.49964

14

0.01561

0.0396494

0.38509

0.97813

0.36948

0.93848

15

0.00460

0.011684

0.09188

0.23338

0.08728

0.22169

16

0.01279

0.0324866

0.18149

0.46098

0.16870

0.42850

17

0.02281

0.0579374

0.34866

0.88560

0.32585

0.82766

18

0.01261

0.0320294

0.13158

0.33421

0.11897

0.30218

19

0.02272

0.0577088

0.11047

0.28059

0.08775

0.22289

20

0.03162

0.0803148

0.22727

0.57727

0.19565

0.49695

21

0.01991

0.0505714

0.13208

0.33548

0.11217

0.28491

30% Polyalkylene Glycol at 59 °F Sample 1

 

1

0.13112

0.3330448

0.13084

0.3323336

−0.00028

−0.0007112

2

0.13057

0.3316478

0.15693

0.3986022

0.02636

0.0669544

3

0.13219

0.3357626

0.19681

0.4998974

0.06462

0.1641348

4

7.20343

18.2967122

7.28350

18.50009

0.08007

0.2033778

5

7.20480

18.300192

7.29568

18.5310272

0.09088

0.2308352

6

7.20677

18.3051958

7.29398

18.5267092

0.08721

0.2215134

7

7.18095

18.239613

7.02154

17.8347116

−0.15941

−0.4049014

8

7.17984

18.2367936

7.07146

17.9615084

−0.10838

−0.2752852

9

7.17900

18.23466

7.13017

18.1106318

−0.04883

−0.1240282

10

0.13277

0.3372358

0.13222

0.3358388

−0.00055

−0.001397

11

0.13222

0.3358388

0.14081

0.3576574

0.00859

0.0218186

12

0.13187

0.3349498

0.13307

0.3379978

0.00120

0.003048

13

0.00322

0.0081788

0.10489

0.2664206

0.10167

0.2582418

14

0.01840

0.046736

0.29129

0.7398766

0.27289

0.6931406

15

0.00662

0.0168148

0.05394

0.1370076

0.04732

0.1201928

16

0.01488

0.0377952

0.07572

0.1923288

0.06084

0.1545336

17

0.02702

0.0686308

0.27715

0.703961

0.25013

0.6353302

18

0.01516

0.0385064

0.15915

0.404241

0.14399

0.3657346

19

0.02298

0.0583692

0.08019

0.2036826

0.05721

0.1453134

20

0.03565

0.090551

0.22438

0.5699252

0.18873

0.4793742

21

0.02236

0.0567944

0.25424

0.6457696

0.23188

0.5889752

Large Depth L Sections

For the large depth L section, Fig. 3, points 1-6 are located on the inner part of the bottom flange and on the top of the web depth along the length. The heights of these points were measured along the depth of the web. Since it is not possible to place the L sections in a configuration similar to Fig. 2(b), accordingly the angle between the flange and the web was measured at three locations, A, B, C, 2.54 cm (1 in), from either edge and at the center.
https://static-content.springer.com/image/art%3A10.1007%2Fs11665-007-9044-0/MediaObjects/11665_2007_9044_Fig3_HTML.gif
Fig. 3

Tall L-section profile

The distortion for two samples is shown in Table 2. The difference between the pre-measure and re-measure values is in hundredths of an inch for water-quenched samples and in thousands of an inch for glycol quenched ones. The average angular distortion was greater than 5° in water and less than 2.5° in glycol.
Table 2

The distortion in cms and in inches for water and glycol quenched large depth L sections

Quenchant

Point

Premeasure, in.

Premeasure, cm

Remeasure, in.

Remeasure, cm

Difference, in.

Difference, cm

Water at 59 °F Sample 1*

 

1

0.15459

0.3926586

0.15308

0.3888232

−0.00151

−0.0038354

2

0.15142

0.3846068

0.13354

0.3391916

−0.01788

−0.0454152

3

0.14980

0.380492

0.15939

0.4048506

0.00959

0.0243586

4

6.85485

17.411319

6.85045

17.400143

−0.00440

−0.011176

5

6.85727

17.4174658

6.84919

17.3969426

−0.00808

−0.0205232

6

6.86108

17.4271432

6.83965

17.372711

−0.02143

−0.0544322

A(°)

89.31098°

85.31082°

−4.00016°

B(°)

89.66091°

88.65220°

−1.00871°

C(°)

89.49681°

95.24684°

5.75003°

30% Polyalkylene Glycol at 59 °F Sample 1*

 

1

0.13448

0.3415792

0.14055

0.356997

0.00607

0.0154178

2

0.14062

0.3571748

0.13804

0.3506216

−0.00258

−0.0065532

3

0.14658

0.3723132

0.13710

0.348234

−0.00948

−0.0240792

4

6.85738

17.4177452

6.85267

17.4057818

−0.00471

−0.0119634

5

6.85530

17.412462

6.84788

17.3936152

−0.00742

−0.0188468

6

6.85432

17.4099728

6.84047

17.3747938

−0.01385

−0.035179

A(°)

90.17744°

91.05474°

0.87730°

B(°)

89.85138°

89.73292°

−0.11846°

C(°)

90.16315°

87.61412°

−2.54903°

* Note that the last 3 (A, B and C) are angles between the web and the flange.

Large Depth T Sections

For the large depth T sections, Fig. 4, points 1-3 and 7-9 are located on the inner part of the flange and 4-6 on top of the web depth along the length. Three angles A, B, C were also recorded. The distortion, Table 3 in glycol-quenched samples is slightly less than that in the water quenched ones. The average distortion was less than 6° in water and less than 4°in glycol.
https://static-content.springer.com/image/art%3A10.1007%2Fs11665-007-9044-0/MediaObjects/11665_2007_9044_Fig4_HTML.gif
Fig. 4

Tall T-section profile

Table 3

The distortion in cms and in inches for water and glycol quenched large depth T sections

Quenchant

Point

Premeasure, in.

Premeasure, cm

Remeasure, in.

Remeasure, cm

Difference, in.

Difference, cm

Water at 59 °F Sample 1

 

1

0.13180

0.334772

0.14120

0.358648

0.00940

0.023876

2

0.13188

0.3349752

0.13861

0.3520694

0.00673

0.0170942

3

0.13944

0.3541776

0.21440

0.544576

0.07496

0.1903984

4

6.85472

17.4109888

6.84069

17.3753526

−0.01403

−0.0356362

5

6.86348

17.4332392

6.87424

17.4605696

0.01076

0.0273304

6

6.86997

17.4497238

6.88828

17.4962312

0.01831

0.0465074

7

0.13246

0.3364484

0.14076

0.3575304

0.00830

0.021082

8

0.13000

0.3302

0.18390

0.467106

0.05390

0.136906

9

0.13472

0.3421888

0.13533

0.3437382

0.00061

0.0015494

A(°)

90.01228

84.80136

−5.21092

B(°)

90.13925

95.27665

5.13740

C(°)

89.56570

90.56397

0.99827

30% Polyalkylene Glycol at 59 °F Sample 1

 

1

0.12987

0.3298698

0.13248

0.3364992

0.00261

0.0066294

2

0.12854

0.3264916

0.13581

0.3449574

0.00727

0.0184658

3

0.13026

0.3308604

0.13141

0.3337814

0.00115

0.002921

4

6.83452

17.3596808

6.82113

17.3256702

−0.01339

−0.0340106

5

6.84977

17.3984158

6.84161

17.3776894

−0.00816

−0.0207264

6

6.86318

17.4324772

6.86205

17.429607

−0.00113

−0.0028702

7

0.13103

0.3328162

0.13597

0.3453638

0.00494

0.0125476

8

0.13202

0.3353308

0.13381

0.3398774

0.00179

0.0045466

9

0.13015

0.330581

0.13094

0.3325876

0.00079

0.0020066

A(°)

89.65550

90.70589

1.05039

B(°)

89.88586

86.02633

−3.85953

C(°)

89.79032

88.65778

−1.13254

Short Depth L and T Sections

For the short depth L and T sections, the web was half (9 cms or 3.5 inches) that of the large depth ones. In Tables 4 and 5 the same number of points was measured as in Tables 2 and 3. The distortion values were minimal (in thousandths of an inch) for both water and glycol quenched samples. Accordingly only one angle was measured between the web and flange and the distortion was almost always within one degree.
Table 4

The distortion in cms and in inches for water and glycol quenched short depth L sections

Quenchant

Point

Premeasure, in.

Premeasure, cm

Remeasure, in.

Remeasure, cm

Difference, in.

Difference, cm

Water at 59 °F Sample 1

 

1

0.13423

0.3409442

0.14177

0.3600958

0.00754

0.0191516

2

0.13171

0.3345434

0.13814

0.3508756

0.00643

0.0163322

3

0.13083

0.3323082

0.13887

0.3527298

0.00804

0.0204216

4

3.50504

8.9028016

3.50122

8.8930988

−0.00382

−0.0097028

5

3.50879

8.9123266

3.50859

8.9118186

−0.00020

−0.000508

6

3.51451

8.9268554

3.50981

8.9149174

−0.00470

−0.011938

Angle(°)

90.39710

89.64311

−0.75399

30% Polyalkylene Glycol at 59 °F Sample 1

 

1

0.13199

0.3352546

0.13158

0.3342132

−0.00041

−0.0010414

2

0.14321

0.3637534

0.14040

0.356616

−0.00281

−0.0071374

3

0.14841

0.3769614

0.14598

0.3707892

−0.00243

−0.0061722

4

3.50777

8.9097358

3.50076

8.8919304

−0.00701

−0.0178054

5

3.51254

8.9218516

3.50732

8.9085928

−0.00522

−0.0132588

6

3.51223

8.9210642

3.51090

8.917686

−0.00133

−0.0033782

Angle(°)

89.74887

89.73433

−0.01454

Table 5

The distortion in cms and in inches for water and glycol quenched short depth T sections

Quenchant

Point

Premeasure, in.

Premeasure, cm

Remeasure, in.

Remeasure, cm

Difference, in.

Difference, cm

Water at 59 °F Sample 1

 

1

0.13108

0.3329432

0.12937

0.3285998

−0.00171

−0.0043434

2

0.13097

0.3326638

0.13169

0.3344926

0.00072

0.0018288

3

0.13146

0.3339084

0.13058

0.3316732

−0.00088

−0.0022352

4

0.13200

0.33528

0.12825

0.325755

−0.00375

−0.009525

5

0.12996

0.3300984

0.12935

0.328549

−0.00061

−0.0015494

6

0.13262

0.3368548

0.13094

0.3325876

−0.00168

−0.0042672

7

3.48887

8.8617298

3.48277

8.8462358

−0.00610

−0.015494

8

3.48792

8.8593168

3.48839

8.8605106

0.00047

0.0011938

9

3.48775

8.858885

3.48787

8.8591898

0.00012

0.0003048

Angle(°)

90.70199

90.36177

−0.34022

30% Polyalkylene Glycol at 59 °F Sample 1

 

1

0.12895

0.327533

0.13507

0.3430778

0.00612

0.0155448

2

0.12959

0.3291586

0.12882

0.3272028

−0.00077

−0.0019558

3

0.13191

0.3350514

0.13023

0.3307842

−0.00168

−0.0042672

4

0.13106

0.3328924

0.13296

0.3377184

0.00190

0.004826

5

0.13014

0.3305556

0.12919

0.3281426

−0.00095

−0.002413

6

0.13191

0.3350514

0.13056

0.3316224

−0.00135

−0.003429

7

3.48685

8.856599

3.48905

8.862187

0.00220

0.005588

8

3.48685

8.856599

3.48774

8.8588596

0.00089

0.0022606

9

3.48706

8.8571324

3.48615

8.854821

−0.00091

−0.0023114

Angle(°)

90.76914

90.93747

0.16833

In the L and T sections the top part where the upper flange was removed experiences the most distortion, Fig. 1 and 5.
https://static-content.springer.com/image/art%3A10.1007%2Fs11665-007-9044-0/MediaObjects/11665_2007_9044_Fig5_HTML.jpg
Fig. 5

A photograph of water quenched sections on the left side of each pair and glycol quenched on the right side

Summary of the Effects of Section Shape

The distortion of the glycol-quenched samples was less than that in water-quenched samples. For the same wall thickness, the I section had the most pronounced distortion along the Z direction followed by the large depth L and T sections. The flanges are 16% thicker than the walls, the I section has four times and two times more flange area as compared to the L and T sections.

Accordingly, there is more area of varying thickness in the I section as compared to the L and T section. Since warpage was more visible in the web of the sections as opposed to the flange, the short depth L and T sections showed minimal distortion, both quenched in water or glycol.

Conclusions

The results of the study indicate that:

  1. 1.

    As expected, warpage is more significant with increasing severity of quench.

     
  2. 2.

    Along the web depth, warpage is more significant in I sections as compared to L and T sections. In the L and T sections the top part where the upper flange was removed experiences the most distortion.

     
  3. 3.

    Warpage increases with increased depth of the web.

     

Acknowledgment

This work was funded by the National Science Foundation, NSF Grant No. EEC-0353668, Research Experiences for the Undergraduates (REU) site, Ms. Esther Bolding is the program manager. The authors wish to thank Mr. R. Marloth of the Mechanical Engineering Department of Loyola Marymount University for reviewing the manuscript and Ms. Maria Guerrero and Ms. Jacquelyn C. Davis for preparing the manuscript.

Copyright information

© ASM International 2007