Par 4-4 Page 4-5
NON HEAT-TREATABLE ALLOYS HEAT-TREATABLE ALLOYS Temper Designation Definition Temper Designation Definition -H3 Strain-hardened and then stabilized.
Applies to products which are strain-hardened and then stabilized by a low temperature heating to slightly lower their strength and increase ductility.
-T651 -T851 -T6510 -T8510 -T6511 -T8511 Solution heat-treated, stress relieved by stretching to produce a permanent set of 1 to 3 percent, and artificially aged.
-H32 Strain-hardened and then stabilized. Final temper is one-quarter hard. -T652 Solution heat-treated, compressed to produce a permanent set, and then artificially aged.
-H34 Strain-hardened and then stabilized. Final temper is one-half hard. -T8 Solution heat-treated, cold-worked, and then artificially aged.
-H36 Strain-hardened and then stabilized. Final temper is three-quarters hard. -T/4 Solution heat-treated, cold-worked by the flattening or straightening operation, and then artificially aged.
-H38 Strain-hardened and then stabilized. Final temper is full-hard. -T86 Solution heat-treated, cold-worked by reduction of 6 percent, and then artificially aged.
-H112 As fabricated; with specified mechanical property limits. -T9 Solution heat-treated, artificially aged and then cold-worked.
-F For wrought alloys; as fabricated.
No mechanical properties limits.
For cast alloys; as cast.
-T10 Cooled from an elevated temperature shaping process artificially aged and then cold-worked.
-F For wrought alloys; as fabricated.
No mechanical properties limits.
For cast alloys; as cast.
4-5.–4-15. [RESERVED.]
Par 4-16 Page 4-6 SECTION 2. TESTING OF METALS
4-16. HARDNESS TESTING. If the material type is known, a hardness test is a simple way to verify that the part has been properly heat-treated. Hardness testers such as Rockwell, Brinell, and Vickers can be useful to check metals for loss of strength due to exposure to fire or abusive heating. Also, understrength bolts can be found and removed from the replacement part inventory by checking the hardness of the bolt across the hex flats. Although hardness tests are generally considered nondestructive, hardness testing does leave a small pit in the surface;
therefore, hardness tests should not be used on sealing surfaces, fatigue critical parts, load bearing areas, etc., that will be returned to service. These hardness tests provide a convenient means for determining, within reasonable limits, the tensile strength of steel. It has several limitations in that it is not suitable for very soft or very hard steels. Hardness testing of aluminum alloys should be limited to distinguishing between annealed and heat-treated material of the same aluminum alloy. In hardness testing, the thickness and the edge distance of the specimen being tested are two factors that must be considered to avoid distortion of the metal. Several readings should be taken and the results averaged. In general, the higher the tensile strength, the greater its hardness. Common methods of hardness testing are outlined in the following paragraphs. These tests are suitable for determining the tensile properties resulting from the heat treatment of steel.
Care should be taken to have case-hardened, corroded, pitted, decarburized, or otherwise nonuniform surfaces removed to a sufficient depth. Exercise caution not to cold-work, and consequently harden, the steel during removal of the surface.
4-17. ROCKWELL HARDNESS TEST. The Rockwell hardness test is the most common method for determining hardness of ferrous and many nonferrous metals. (See Table 4-5.) It differs from Brinell hardness testing in that the hardness is determined by the depth of indentation made by a constant load impressing on an indenter. In this test, a standard minor load is applied to set a hardened steel ball or a diamond cone in the surface of the metal, followed by the application of a standard major load. The hardness is measured by depth of penetration.
Rockwell superficial hardness tests are made using light minor and major loads and a more sensitive system for measuring depth of indentation. It is useful for thin sections, very small parts, etc. Calibration of Rockwell hardness testers is done in accordance with ASTM International (ASTM) E-18 specifications.
Par 4-17 Page 4-7 TABLE 4-5. HARDNESS VALUES FOR ALUMINUM ALLOYS (refer to AMS2772H) Material Commercial Designation Hardness Temper Brinell Number 500 kg. Load 10 mm. Ball 1100 0 23 H18 44 3003 0 28 H16 47 2014 0 45 T6 135 2017 0 45 T6 105 2024 0 47 T4 120 2025 T6 110 6151 T6 100 5052 0 47 H36 73 6061 0 30 T4 65 T6 95 7075 T6 135 7079 T6 135 195 T6 75 220 T4 75 C355 T6 80 A356 T6 70
4-18. BRINELL HARDNESS TEST. In this test, a standard constant load (usually 500 kg to 3,000 kg) is applied to a smooth flat metal surface by a hardened steel-ball type indenter 10 mm in diameter. The 500-kg load is usually used for testing nonferrous metals such as copper and aluminum alloys, whereas the 3,000-kg load is most often used for testing harder metals such as steels and cast irons. The numerical value of Brinell Hardness (HB) is equal to the load divided by the surface area of the resulting spherical impression.
Wh ere:
P=load in kg D=diameter of the ball in mm d=diameter of the indentation in mm
Par 4-18 Page 4-8
a. General Precautions. To avoid misapplication of Brinell hardness testing, the fundamentals and limitations of the test procedure must be clearly understood. To avoid inaccuracies, the following rules should be followed:
(1) Do not make indentations on a curved surface having a radius of less than 1 inch.
(2) Do make the indentations with the correct spacing. Indentations should not be made too close to the edge of the work piece being tested.
(3) Apply the load steadily to avoid overloading caused by inertia of the weights.
(4) Apply the load so the direction of loading and the test surface are perpendicular to each other within 2 degrees.
(5) The thickness of the work piece being tested should be such that no bulge or mark showing the effect of the load appears on the side of the work piece opposite the indentation.
(6) The indentation diameter should be clearly outlined.
b. Limitations. The Brinell hardness test has three principal limitations:
(1) The work piece must be capable of accommodating the relatively large indentations.
(2) Due to the relatively large indentations, the work piece should not be used after testing.
(3) The limit of hardness, 15 HB with the 500-kg load to 627 HB with the 3,000-kg load, is generally considered the practical range.
c. Calibration. A Brinell hardness tester should be calibrated to meet ASTM standard E10 specifications.
4-19. VICKERS HARDNESS TEST. In this test, a small pyramidal diamond is pressed into the metal being tested. The Vickers Hardness number (HV) is the ratio of the load applied to the surface area of the indention. This is done with the following formula:
a.
The indenter is made of diamond, and is in the form of a square-based pyramid having an angle of 136 degrees between faces. The facets are highly polished, free from surface imperfections, and the point is sharp. The loads applied vary from 1 to 120 kg; the standard loads are 5, 10, 20, 30, 50, 100, and 120 kg. For most hardness testing, 50 kg is maximum.
b. A Vickers hardness tester should be calibrated to meet ASTM Standard E10 specifications, acceptable for use over a loading range.
Par 4-20 Page 4-9
4-20. MICROHARDNESS TESTING. This is an indentation hardness test made with loads not exceeding 1 kg (1,000 g). Such hardness tests have been made with a load as light as 1 g, although the majority of microhardness tests are made with loads of 100 g to 500 g. In general, the term is related to the size of the indentation rather than to the load applied.
a. Fields of Application. Microhardness testing is capable of providing information regarding the hardness characteristics of materials which cannot be obtained by hardness tests such as the Brinell or Rockwell, and are as follows:
(1) Measuring the hardness of precision work pieces that are too small to be measured by the more common hardness testing methods.
(2) Measuring the hardness of product forms such as foil or wire that are too thin or too small in diameter to be measured by the more conventional methods.
(3) Monitoring of carburizing or nitriding operations, which is sometimes accomplished by hardness surveys taken on cross-sections of test pieces that accompanied the work pieces through production operations.
(4) Measuring the hardness of individual microconstituents.
(5) Measuring the hardness close to edges, thus detecting undesirable surface conditions such as grinding burn and decarburization.
(6) Measuring the hardness of surface layers such as plating or bonded layers.
b. Indenters. Microhardness testing can be performed with either the Knoop or the Vickers indenter. The Knoop indenter is used mostly in the United States; the Vickers indenter is more widely used in Europe.
(1) Knoop indentation testing is performed with a diamond, ground to pyramidal form, that produces a diamond-shaped indentation with an approximate ratio between long and short diagonals of 7 to 1. The indentation depth is about one-thirtieth of its length. Due to the shape of the indenter, indentations of accurately measurable length are obtained with light loads.
(2) The Knoop Hardness number (HK) is the ratio of the load applied to the indenter to the unrecovered projected area of indentation. The formula for this follows:
Wh ere:
P=the applied load in kg A=the unrecovered projected area in square mm l=the measured length of the long diagonal in mm C=0.07028 (a constant of the indenter, relating projected area of the indentation to the square of the length of the long diagonal)
Par 4-21 Page 4-10
4-21. INDENTATIONS. The Vickers indenter penetrates about twice as far into the work piece as does the Knoop indenter. The diagonal of the Vickers indentation is about one-third of the total length of the Knoop indentation. The Vickers indenter is less sensitive to minute differences in surface conditions than is the Knoop indenter. However, the Vickers indentation, because of the shorter diagonal, is more sensitive to errors in measuring than is the Knoop indentation. (See Figure 4-1.)
FIGURE 4-1. COMPARISON OF INDENTATION MADE BY KNOOP AND VICKERS
INDENTERS IN THE SAME WORK METAL AND AT THE SAME LOADS 4 -22. MAGNETIC TESTING. Magnetic testing consists of determining whether the specimen is attracted by a magnet. Usually, a metal attracted by a magnet is iron, steel, or an iron-base alloy containing nickel, cobalt, or chromium. However, there are exceptions to this rule since some nickel and cobalt alloys may be either magnetic or nonmagnetic. Never use this test as a final basis for identification. The strongly attracted metals could be pure iron, pure nickel, cobalt, or iron-nickel-cobalt alloys. The lightly attracted metals could be cold-worked stainless steel or Monel. The nonmagnetic metals could be aluminum, magnesium, silver, or copper-base alloy, or an annealed 300-type stainless steel.
4-23. ALUMINUM TESTING. Hardness tests are useful for testing aluminum alloy chiefly as a means of distinguishing between annealed, cold-worked, heat-treated, and heat-treated and aged material. It is of little value in indicating the strength or quality of heat-treatment. Typical hardness values for aluminum alloys are shown in Table 4-5.
a. Clad aluminum alloys have surface layers of pure aluminum or corrosion-resistant aluminum alloy bonded to the core material to inhibit corrosion. Presence of such a coating may be determined under a magnifying glass by examination of the edge surface which will show three distinct layers. In aluminum alloys, the properties of any specific alloy can be altered by work hardening (often called strain hardening), heat treatment, or by a combination of these processes.
Par 4-23 Page 4-11
b. Test for distinguishing heat-treatable and nonheat-treatable aluminum alloys. If for any reason the identification mark of the alloy is not on the material, it is possible to distinguish between some heat-treatable alloys and some nonheat-treatable alloys by immersing a sample of the material in a 10 percent solution of caustic soda (sodium hydroxide). Those heat-treated alloys containing several percent of copper (2014, 2017, and 2024) will turn black due to the copper content. High-copper alloys when clad will not turn black on the surface, but the edges will turn black at the center of the sheet where the core is exposed. If the alloy does not turn black in the caustic soda solution it is not evidence that the alloy is nonheat-treatable, as various high-strength heat-treatable alloys are not based primarily on the use of copper as an alloying agent. These include, among others, 6053, 6061, and 7075 alloys. The composition and heat-treating ability of alloys which do not turn black in a caustic soda solution can be established only by chemical or spectro-analysis.
4-24.–4-35. [RESERVED.]
Par 4-36 Page 4-12 SECTION 3. PRECAUTIONARY MEASURES
4-36. FLUTTER AND VIBRATION PRECAUTIONS. To prevent the occurrence of severe vibration or flutter of flight control surfaces during flight, precautions must be taken to stay within the design balance limitations when performing maintenance or repair.
a. Balance Changes. The importance of retaining the proper balance and rigidity of aircraft control surfaces cannot be overemphasized. The effect of repair or weight change on the balance and center of gravity is proportionately greater on lighter surfaces than on the older heavier designs. As a general rule, repair the control surface in such a manner that the weight distribution is not affected in any way, in order to preclude the occurrence of flutter of the control surface in flight. Under certain conditions, counter-balance weight is added forward of the hinge line to maintain balance. Add or remove balance weights only when necessary in accordance with the manufacturer’s instructions. Flight testing must be accomplished to ensure flutter is not a problem. Failure to check and retain control surface balance within the original or maximum allowable value could result in a serious flight hazard.
b. Painting and Refinishing. Special emphasis is directed to the effect of too many extra coats of paint on balanced control surfaces. Mechanics must avoid adding additional coats of paint in excess of what the manufacturer originally applied. If available, consult the aircraft manufacturer’s instructions relative to finishing and balance of control surfaces.
c. Trapped Water or Ice. Instances of flutter have occurred from unbalanced conditions caused by the collection of water or ice within the surface. Therefore, ventilation and drainage provisions must be checked and retained when maintenance is being done.
d. Trim Tab Maintenance. Loose or vibrating trim tabs will increase wear of actuating mechanisms and hinge points which may develop into serious flutter conditions. When this happens, primary control surfaces are highly susceptible to wear, deformation, and fatigue failures because of the buffeting nature of the airflow over the tab mechanism. Trailing-edge play of the tab may increase, creating an unsafe flutter condition. Careful inspection of the tab and its mechanism should be conducted during overhaul and annual inspection periods.
Compared to other flight control systems on the aircraft, only a minor amount of tab-mechanism wear can be tolerated.
(1) Free play and stiffness may best be measured by a simple static test where “upward” and “downward” (or “leftward” and “rightward”) point forces are applied near the trailing edge of the tab at the span-wise attachment of the actuator (so as not to twist the tab). The control surface to which the trim tab is attached should be locked in place. Rotational deflection readings are then taken near the tab trailing edge using an appropriate measuring device, such as a dial gauge. Several deflection readings should be taken using loads first applied in one direction, then in the opposite. If the tab span does not exceed 35 percent of the span of the supporting control surface, the total free play at the tab trailing edge should not exceed 2 percent of the tab chord. If the tab span equals or exceeds 35 percent of the span of the supporting control surface, the total free play at the tab trailing edge should not exceed 1 percent of the distance from the tab hinge line to the trailing edge of the tab perpendicular to the tab hinge line. For example, a tab that has a chord of 4 inches and less than or equal to 35 percent of the control surface span would have a
Par 4-36 Page 4-13 maximum permissible free play of 4 inches x 0.020 or 0.080 inches (total motion up and down) measured at the trailing edge. Correct any free play in excess of this amount.
(2) Care must also be exercised during repair or rework to prevent stress concentration points or areas that could increase the fatigue susceptibility of the trim tab system. Advisory Circular (AC) 23.629-1, Means of Compliance with Title 14 CFR, Part 23, § 23.629, Flutter, contains additional information on this subject.
NOTE: If the pilot has experienced flutter, or thinks they have, then a complete inspection of the aircraft flight control system and all related components, including rod ends, bearings, hinges, and bellcranks, must be accomplished. Suspected parts should be replaced.
4-37. LOAD FACTORS FOR REPAIRS. In order to design an effective repair to a sheet metal aircraft, the stresses that act on the structure must be understood.
a. Six types of major stresses are known and should be considered when making repairs.
These are tension, compression, bending, torsion, shear, and bearing.
b. The design of an aircraft repair is complicated by the requirement that it be as light as possible. If weight were not critical, repairs could be made with a large margin of safety. But in actual practice, repairs must be strong enough to carry all of the loads with the required factor of safety, but they must not have too much extra strength. A joint that is too weak cannot be tolerated, but neither can one that is too strong because it can create stress risers that may cause cracks in other locations.
4-38. TRANSFER OF STRESSES WITHIN A STRUCTURE. An aircraft structure must be designed in such a way that it will accept all of the stresses imposed upon it by the flight and ground loads without any permanent deformation. Any repair made must accept the stresses, carry them across the repair, and then transfer them back into the original structure. These stresses are considered as flowing through the structure, so there must be a continuous path for them, with no abrupt changes in cross-sectional areas along the way. Abrupt changes in cross-sectional areas of aircraft structure that are subject to cycle loading/stresses will result in stress concentration that may induce fatigue cracking and eventual failure. A scratch or gouge in the surface of a highly stressed piece of metal will cause a stress concentration at the point of damage.
a. Multirow Fastener Load Transfer. When multiple rows of rivets are used to secure a lap joint, the transfer of stresses is not equal in each row. The transfer of stress at each row of rivets may be thought of as transferring the maximum amount capable of being transferred without experiencing rivet shear failure.
b. Use Of Stacked Doublers. A stacked doubler is composed of two or more sheets of material that are used in lieu of a single, thicker sheet of material. Because the stress transferred at each row of rivets is dependent upon the maximum stress that can be transferred by the rivets in that row, the thickness of the sheet material at that row need only be thick enough to transfer the stress applied. Employing this principle can reduce the weight of a repair joint.
4-39.–4-49. [RESERVED.]
Par 4-50 Page 4-14 SECTION 4. METAL REPAIR PROCEDURES
4-50. GENERAL. The airframe of a fixed-wing aircraft is generally considered to consist of five principal units: the fuselage, wings, stabilizers, flight control surfaces, and landing gear.
a. Aircraft Principal Structural Elements (PSE) and joints are designed to carry loads by distributing them as stresses. The elements and joints as originally fabricated are strong enough to resist these stresses, and must remain so after any repairs. Long, thin elements are called members. Some examples of members are the metal tubes that form engine mount and fuselage trusses and frames, beams used as wing spars, and longerons and stringers of metal-skinned fuselages and wings. Longerons and stringers are designed to carry principally axial loads, but are sometimes required to carry side loads and bending moments, as when they frame cutouts in metal-skinned structures. Truss members are designed to carry axial (tension and compression) loads applied to their ends only. Frame members are designed to carry side loads and bending moments in addition to axial loads. Beam members are designed to carry side loads and bending moments that are usually large compared to their axial loads. Beams that must resist large axial loads, particularly compression loads, in combination with side loads and bending moments are called beam-columns. Other structural elements such as metal skins, plates, shells, wing ribs, bulkheads, ring frames, intercostal members, gussets, other reinforcements, and fittings are designed to resist complex stresses, sometimes in three dimensions.
b. Any repair made on an aircraft structure must allow all of the stresses to enter, sustain these stresses, and then allow them to return into the structure. The repair must be equal to the original structure, but not stronger or stiffer, which will cause stress concentrations or alter the resonant frequency of the structure.
c. All-metal aircraft are made of very thin sheet metal, and it is possible to restore the strength of the skin without restoring its rigidity. All repairs should be made using the same type and thickness of material that was used in the original structure. If the original skin had corrugations or flanges for rigidity, these must be preserved and strengthened. If a flange or corrugation is dented or cracked, the material loses much of its rigidity; and it must be repaired in such a way that will restore its rigidity, stiffness, and strength.
4-51. RIVETED (OR BOLTED) STEEL TRUSS-TYPE STRUCTURES. Repairs to riveted structures may be made employing the general principles outlined in the following paragraphs on aluminum alloy structures. Repair methods may also be found in textbooks on metal structures.
Methods for repair of the major structural members must be specifically approved by the Federal Aviation Administration (FAA).
4-52. ALUMINUM ALLOY STRUCTURES. Extensive repairs to damaged stressed skin on monocoque-types of aluminum alloy structures must be made in accordance with FAA-approved manufacturer’s instructions or another FAA-approved source.
a. Rivet Holes. Rivet holes are slightly larger than the diameter of the rivet. When driven, solid rivets expand to fill the hole. The strength of a riveted joint is based upon the expanded
Par 4-52 Page 4-15 diameter of the rivet. Therefore, it is important that the proper drill size be used for each rivet diameter.
(1) The acceptable drill size for rivets may be found in MIL-HDBK-5, Metallic Materials and Elements for Aerospace Vehicle Structures, or MMPDS-10.
(2) Avoid drilling oversized holes or otherwise decreasing the effective tensile areas of wing-spar cap strips, wing, fuselage, fin longitudinal stringers, or highly stressed tensile members. Make all repairs, or reinforcements, to such members in accordance with factory recommendations or with the specific approval of an FAA representative.
b. Disassembly Prior to Repairing. If the parts to be removed are essential to the rigidity of the complete structure, support the structure prior to disassembly in such a manner as to prevent distortion and permanent damage to the remainder of the structure. When rivets are removed, undercut rivet heads by drilling. Use a drill of the same size as the diameter of the rivet. Drilling must be exactly centered and to the base of the head only. After drilling, break off the head with a pin punch and carefully drive out the shank. On thin or unsupported metal skin, support the sheet metal on the inside with a bucking bar. Removal of rivet heads with a cold chisel and hammer is not recommended because skin damage and distorted rivet holes will probably result. Inspect rivet joints adjacent to damaged structure for partial failure by removing one or more rivets to see if holes are elongated or the rivets have started to shear.
c. Effective Tools. Care must also be taken whenever screws must be removed to avoid damage to adjoining structure. When properly used, impact wrenches can be effective tools for removal of screws; however, damage to adjoining structure may result from excessive vertical loads applied through the screw axis. Excessive loads are usually related to improperly adjusted impact tools or attempting to remove screws that have seized from corrosion. Remove seized screws by drilling and use of a screw extractor. Once the screw has been removed, check for structural cracks that may appear in the adjoining skin doubler, or in the nut or anchor plate.
4-53. SELECTION OF ALUMINUM FOR REPLACEMENT PARTS. All aluminum replacement sheet metal must be identical to the original or properly altered skin. If another alloy is being considered, refer to the information on the comparative strength properties of aluminum alloys contained in MIL-HDBK-5 or MMPDS-10.
a. Temper. The choice of temper depends upon the severity of the subsequent forming operations. Parts having single curvature and straight bend lines with a large bend radius may be advantageously formed from heat-treated material; while a part, such as a fuselage frame, would have to be formed from a soft, annealed sheet, and heat-treated after forming. Make sure sheet metal parts which are to be left unpainted are made of clad (aluminum coated) material. Make sure all sheet material and finished parts are free from cracks, scratches, kinks, tool marks, corrosion pits, and other defects which may be factors in subsequent failure.
b. Use of Annealed Alloys for Structural Parts. The use of annealed aluminum alloys for structural repair of an aircraft is not recommended. An equivalent strength repair using annealed aluminum will weigh more than a repair using heat-treated aluminum alloy.
Par 4-54 Page 4-16
4-54. HEAT TREATMENT OF ALUMINUM ALLOY PARTS. All structural aluminum alloy parts are to be heat-treated in accordance with the heat-treatment instruction issued by the manufacturers of the part. In the case of a specified temper, the sequence of heat-treating operations set forth in MIL-HDBK-5 or MMPDS-10 and corresponding specifications. If the heat-treatment produces warping, straighten the parts immediately after quenching. Heat-treat riveted parts before riveting, to preclude warping and corrosion.
a. Quenching. Quench material from the solution heat-treating temperature as rapidly as possible after removal from the furnace. Quenching in cold water is preferred, although less drastic chilling (hot or boiling water, or airblast) is sometimes employed for bulk sections, such as forgings, to minimize quenching stresses.
b. Reheating at Temperatures Above Boiling Water. Reheating of 2017 and 2024 alloys above 212 °F tend to impair the original heat treatment. Therefore, reheating above 212 °F, including the baking of primers, is not acceptable without subsequent complete and correct heat treatment.
4-55. BENDING METAL. When describing a bend in aviation, the term “bend radii” (BR) is used to refer to the inside radius. Requirements for bending the metal to various shapes are frequently encountered. When a metal is bent, it is subjected to changes in its grain structure, causing an increase in its hardness.
a. The minimum radius is determined by the composition of the metal, its temper, and thickness. Table 4-6 shows the recommended radius for different types of aluminum. Note that the smaller the thickness of the material, the smaller the recommended minimum BR, and that as the material increases in hardness, the recommended BR increases.
b. When using layout techniques, the mechanic must be able to calculate exactly how much material will be required for the bend. It is easier to lay out the part on a flat sheet before the bending or shaping is performed. Before bending, smooth all rough edges, remove burrs, and drill relief holes at the ends of bend lines and at corners to prevent cracks from starting. Bend lines should preferably be made to lie at an angle to the grain of the metal (preferably 90 degrees).
c. BR in inches for a specific metal composition (alloy) and temper is determined from Table 4-6. For example, the minimum BR for 0.016 thick 2024-T6 (alloy and temper) is found is found to be 2 to 4 times the material thickness, or 0.032 to 0.064.
4-56. SETBACK.
a. Setback is a measurement used in sheet metal layout. It is the distance the jaws of a brake must be setback from the mold line to form a bend. For a 90-degree bend, the point is back from the mold line to a distance equal to the BR plus the metal thickness. The mold line is an extension of the flat side of a part beyond the radius. The mold line dimension of a part is the dimension made to the intersection of mold lines, and is the dimension the part would have if its corners had no radius. (See Figure 4-2.)