二輥棒材矯直機的設計含5張CAD圖

二輥棒材矯直機的設計含5張CAD圖,二輥棒材,矯直機,設計,CAD ORIGINAL ARTICLEResearch and verification on neutral layer offset of bar in two-rollstraightening processLifeng Ma&Ziyong Ma&Weitao Jia&Yangyang Lv&Yaping Jiang&Haijie Xu&Pengtao LiuReceived: 3 August 2014 /Accepted: 9 February 2015 /Published online: 6 March 2015# Springer-Verlag London 2015Abstract Aneutrallayermustbeexistedinthemetalmaterialdeformation. Traditional two-roll straightening theory ignoresthe migration phenomenon of neutral layer, especially forlarge cross-sectional bar in the straightening process. Neutrallayer has a large impact on straightening springback and thenaffects the roller shape design, process parameters, andstraightness accuracy of bar. In the present work, a neutrallayer offset model was established in the bar straighteningprocess based on the three-point bending and elastic-plasticpressure. The neutral layer migration phenomenon had beenstudied through the model in the bar straightening process,combinedwithroom-temperaturetensiletestandbendingtest.The relationship of neutral layer offset values and reversebending radius and plastic deformation capacity of the metalhas been obtained. The neutral layer offset model provides areference to further study of bar straightening mechanism anddeformation.Keywords Neutrallayeroffset.Barstraightening.Springback.Three-pointbending.Experimentalanalysis1 IntroductionHigh-strength alloy steel bars are increasingly used in theareas of petroleum, automobile, shipbuilding, and construc-tion machinery. As the final finishing step, straightening isthe key process to ensure the quality of bar produced 1.Because original bending of bar is likely to exist in any direc-tion, it is particularly suitable for cross-roll straightener. Two-roll straightening and multi-roll straightening are two mainforms of cross-roll straightener. However, the existing multi-roll straightening machine has many problems, such as bigblind area of straightening, disability of full-length and full-continuous straightening, and need of removing the head andend after straightening. Moreover, the complex structure andlarge volume of multi-roll straightener generally cause highcost. Because of low precision of straightening, multi-rollstraightener cannot realize the high precision of bar straight-ening. Compared to the multi-roll straightener, two-rollstraightening machine cannot only eliminate the blind areaof bar and improve surface roughness, but also improve oval-ityandsolveneckingafterstraightening.Moreimportantly,itsresidual deflection of bar can achieve 0.10.5 mm/m afterstraightening, which meets the current accuracy requirementsof high precision bar. Domestic small two-roll straighteningdevice has been manufactured, but it has been failed to masterthe roller design, and core technology of automatic straight-ening process, and the two-roll straightening machine of largecross-sectional bar entirely depends on import. Therefore,based on the localization of device, our research group con-ducted a collaborative work with a company in HebeiProvince, China, to develop the high precision two-roll barstraightening machine. The deviation and establishment ofneutral layer offset model are problems 2, 3, which are di-rectly related to the precision of the roll shape design andstraightening accuracy.Two-roll straightening is a complex process of elastic-plastic deformation. Neutral layer offset changes stress distri-bution of cross section of bar, affects the bending momentratio and the calculated reverse bend curvature, finally affectsthe accuracy of roll shape design. Unfortunately, the existingtheoretical analysis on bar straightening has ignored the phe-nomenon of neutral layer offset; therefore, the designed rollL. Ma:Z. Ma (*):W. Jia:Y. Lv:Y. Jiang:H. Xu:P. LiuHeavy Machinery Engineering Research Center of MinistryEducation, Taiyuan University of Science and Technology,Taiyuan 030024, Chinae-mail: Z. Mae-mail: Int J Adv Manuf Technol (2015) 79:15191529DOI 10.1007/s00170-015-6899-3shape and process parameters have large error. The study ofGuan et al. showed that under certain relative corner radius,therelativeerror inthe neutral layerbending springbackcouldbe up to 70 % or more 4. At present, for the issue on theneutral layer offset, many scholars have done lots of research58. Their research has focused on plate and pipe instead ofthe neutral layer offset in the two-roll straightening process.So based on the characteristics of its straightening process,combined with three-point bending theory and elastic-plastictheory, a neutral layer offset model is established in thestraightening process. This work provides an important refer-enceforfurtherstudyonthetwo-rollstraighteningmechanismof large cross section and high-strength steel bars and devicedevelopment.2 Establishment of the theoretical model of neutral layeroffset in two-roll straightening2.1 Straightening deformation analysis and basic assumptions1.Cross section of bar remains flat before and after defor-mation is perpendicular to the axis of the deformed bar9. There are no reverse and tilt between two adjacentcross sections.2.The material is continuous, homogeneous, and isotropic,and the stressand strainofneutral layersare consideredtobe coincident 10.3.The diameter change during bar straightening process isignored.4.The bending plastic deformation process is consistentwith the principle of constant volume.5.Equivalent stress and equivalent strain have a harden-ing exponent relationship of Bn, where B is the co-efficient ofplastic materialand n isthe materialhardeningexponent 11.When two-roll straightener straightens bars, bending de-flection is determined by concave roll angle and roll gap. Byadjusting the roll gap, the actual bending deflection can varyfrom 0 to the maximum bending deflection 12. A more rea-sonable straightening state is shown in Fig. 1; at this point,straightening deformation is approximate to three-pointbending.A unit of ABCD is taken from the bar straightening defor-mation zone, as shown in Fig. 2. In the figure, bar bendingcenter is the original point of cylindrical coordinate system.According to hypothesis (3), deformation in r direction isnegligible, i.e., r=0. direction stress is much smaller thanothersandcanbeignored,namely=0.Therefore,theunitsstress state of bar straightening deformation zone can be sim-plified toplane strain problem, i.e., = r= 0 and = r= 0. Meanwhile, the hypothesis (1) means r=r=0.2.2 Strain relation in straightening processBeforeentering straightener, the originalforms ofbar bendingcausedbymachiningorheattreatmenthavesingle,Sbending,multi-peak, and space bending types 13 in the length range,as shown in Fig. 3. All of them will be unified into a singlebending type in the central section of straightening roller.Generally, the original bending form is simplified as singlebending type to facilitate the theoretical analysis.It can be assumed that the unit ABCD is initially in tensilestateandinthecompressionstateafterreversebending14.Theentering straightener bar has small degree of bending. The neu-tral layer and the geometric center axis of the bar are consideredto be coincident, so the original length of the bar unit l0is:l0 R001where R0is the original bending radius and 0is the originalbending angle of unit.The original fiber length of unit ABCD l0Dis:l0D r002where r0is the original bending radius of unit.After unit being bent, the length of neutral layer lwis:lw w3where isthe radiusofneutral layer after reverse bendingandwis the reverse bending angle. At this point, the fiber lengthof the unit ABCD lwDis:lwD rw4So the true strain of the unit in the tangential direction is: lnlwDl0D lnrwr005The lengthofbarneutrallayerisa constantbeforeandafterstraightening that means:R00 w6Fig. 1 Diagram of actual straightening state1520Int J Adv Manuf Technol (2015) 79:15191529Finally, lnrR0R0 Rw? r7where Rwis the reverse bending radius of bar.2.3 Stress and strain relationship of plastic deformation zoneTheareaofplasticdeformationofthebarmeetstherelevantlawsof plastic deformation. According to hypothesis (4), =.In the plane plastic deformation, r direction has no defor-mation, i.e., dr=0.According to the incremental theory 15:r128So, the equivalent strain pand the equivalent stress pinplastic deformation zone are:pffiffiffi2p3ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi? r2 ? 2 ? r2q2ffiffiffi3pjj9p1ffiffiffi2pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi? r2 ? 2 ? r2qffiffiffi3p2jj102.4 The relationship between external force and plasticdeformation zone stressHalf-space as shown in Fig. 4, C(, ) is taken in the innersurface of loading area S, and A(x, y, z) is a point in the solid16. So the distance is:0 ? x2 ? y2 z2hi1211The potential function is defined as:H1 Sp ;dd12where =zln(0+z)0The potential function is defined as:H H1z Sp ;ln 0 zdd13Fig. 2 Diagram of barstraightening stress and strain indeformation zoneFig. 3 Original bending types ofbarInt J Adv Manuf Technol (2015) 79:151915291521Definition as:1H1z; Hz Sp ;10ddDisplacement function in z direction is:uz14G12 ? zz?14Stress function in z direction is:z12z? z2z2?15Boundary condition is:z?p ;In the sregion0Outside the sregion?When concentrated force P acts on the surface of originvertically, 0 x2 y2 z212, Sp(,)d d=PSimplify defined Boussinesq potential function to:1H1z Pln 0 z HzP016Stress component in z direction is expressed as:z r ?3P2z305172.5 Mathematical model of the neutral layer offsetAn elastic deformation zone exists near neutral layer, in addi-tion to the plastic deformation zone, when bar is in thestraightening process. Such a relationship sis drawn intheir boundary surface that meansffiffiffi3p ?3P21r ? Rwd2?2? s18A conclusion is drawn by hypothesis (5)B2ffiffiffi3p jj?n s19namelyB2ffiffiffi3p lnR0rR0 Rw? r?n s20Eliminate r by simultaneous equations R0eARw?d2ffiffiffiffiffiffiffiffiffiffiffiffiffi3ffiffiffi3pP2Ss01AR0d2?ffiffiffiffiffiffiffiffiffiffiffiffiffi3ffiffiffi3pP2Ss21The half-space assumptions applied in the theoretical deri-vationdiffersfromthecontactconditionofbarandtheconvexroll, so a correction factor is applied R0eARw?d2 T?R0d2? T22where A ffiffi3p2SB? ?1n, T ffiffiffiffiffiffiffiffiffi3ffiffi3pP2Sq, both of which are coeffi-cients related to the material, d is the bar diameter, sisthe yield strength of the bar, B is the plastic hardeningcoefficient, n is the hardening exponent, and is thecorrection factor related with material and bending de-gree and determined by the data fitting of simulationand experiment.Fig. 4 Stress and displacementdiagram of half-spaceconcentration1522Int J Adv Manuf Technol (2015) 79:15191529Therefore, the mathematical model of neutral layer offsetvalue is RwR0eA Rw?d2 T?R0d2? T233 Simulation of neutral layer offset in two-rollstraightening process3.1 3D finite element model of bendingFinite element analysis model developed is shown in Fig. 5.Barisplasmodium,andindenterisassumedasrigidbody.Thebar length is 340 mm, and the original maximum deflection is10 mm/m. Material model is the bilinear kinematic hardeningmodel, and the yield strength, elastic modulus, and Poissonsratio are shown in Table 2. Bar and indenter are meshed byusing hexahedral elements solid164 and sweep 17. All con-tacts are defined as surface-to-surface automatic contact in thedeformation process, the static friction coefficient is 0.25, anddynamic friction coefficient is 0.15.3.2 Simulation resultsFigure6istheplasticdeformationdepthdiagramsof40Crand42CrMo with the same reductions. It can be seen that thedepthofplasticdeformationlayeroftwomaterialsinthe samereductions is different. The ratios of plastic deformation depthand radius of 40Cr and 42CrMo are approximately 0.91 and0.74,respectively,andso,plasticdeformationdepthof40Crisgreaterthanthatof42CrMo,andtheplasticdeformationdepthof tensile side is greater than the pressure side. The depthratios of tensile side and pressure side plastic deformation of40Cr and 42CrMo are 1.077 and 1.094, respectively, whichindirectly proved that the neutral layer will move to the pres-sure side.The bar is divided into two halves in the axial direction.Taking different units inside the bar and obtaining their stresscurves in the tangent direction (X direction), it can be seenfrom the stress curves shown in Fig. 7a that element no.104824 is stretched by tensile stress in 0.0402 s and elementno. 104710 is compressed by compressive stress. At the sametime, the stress neutral layer must be between them. Elementno. 104710 subjects to tensile stress in 0.0804 s, and elementno. 104596 subjects to compressive stress; stress neutral layershifts between them. In 0.1474 s, the stress neutral layer devi-ates to the place between element no. 104596 and element no.104482. Similarly, asshown inFig. 7b, the stressneutral layeris between element nos. 104824 and 104938 in the first place.With the degree of bending increases gradually, it shifts be-tween element nos. 104824 and 104710, element nos. 104710and104596, and elementnos.104596and104824. Therefore,it is possible to determine the value of stress neutral layeroffset relative to the geometric center axis according to theelement stress curves and simulation data of two materials asshown in Table 1.4 Neutral layer offset experiments in the process of tworollers straightening4.1 Experimental route and purposeThe experimental route and purpose are shown in Fig. 8.4.2 Uniaxial tensile tests at room temperatureTensile test is one of the most basic experiments to measurethe mechanical properties of materials. By a tensile test, theyield strength, elongation, tensile strength, and other parame-ters of a material can be obtained, which are important factorsof bar bending deformation 18. In this paper, 40Cr and(a)(b)Fig. 6 Plastic deformation depth diagrams of materials with the samereductions: a 40Cr and b 42CrMoFig. 5 Finite element modelInt J Adv Manuf Technol (2015) 79:15191529152342CrMo are used for tensile test. Experimental data of tensiletest is analyzed, which provides parameters support for theconstruction of a neutral layer offset model.Tensile tests were conducted on a WAW 1000 universaltesting machine. The resulting data had been processed toget the stress and strain curves as shown in Fig. 9.The Origin Software is applied to fitting the tensile exper-imental data. The data is taken from the yield point to tensilestrength. The fitting results show that the correction coeffi-cients of determination fitted by the power function are re-spectively 0.9539 and 0.9647 after 40Cr and 42CrMo enter(a)(b)Fig. 7 Stress change curves ofmaterials in the X direction: a40Cr and b 42CrMoTable 1Simulated data of neutral layer offset under different reversebending radiusMaterialDiameterd (mm)TimeT (s)Reduction (mm)Bending radiusRW(mm)Offset value (mm)40Cr280.0402 5.022875.8680.19530.0678.3751725.8180.42330.0804 10.051438.2640.48840.0938 11.7251232.8840.61860.1206 15.075959.0770.78140.1474 18.425784.8060.911642CrMo 250.0402 5.022875.9020.13020.0678.3751725.1590.22790.0804 10.051437.9090.32560.0938 11.7251231.0330.39070.1206 15.075959.0770.65120.1474 18.425784.6310.7814Fig. 8 Study route and purpose1524Int J Adv Manuf Technol (2015) 79:15191529the plastic deformation. The power function adjusted determi-nationcoefficientsarecloseto1.Sincethefittingresultisverygood, the stress and strain relationship of 40Cr and 42CrMocan be effectively described by the power function hardeningrelationship of Bn. The power function fitting results ofthe two kinds of materials are shown in Table 2.4.3 Bending testsBending testisapplied todetermine the mechanicalpropertieswhen material subjected to bending load. It is one of the basicmethods of mechanical property tests. The bar is simply sup-ported at both ends with a concentrated load applied in themiddle. The load is loaded by 100 kN of hydraulic pressuretesting machine. Hydraulic pressure testing machine is con-trolled by electro-hydraulic servo pressure testing machinemeasurement and control system. The load value can be con-trolled by electro-hydraulic servo testing machine pressuremeasurement and control system.4.3.1 Load-displacement curve of pressure processThe load-displacement curve of pressure process is shown inFig. 10.4.3.2 The post-processing of bending barFirstly, using Nikon S4150 camera was used to take picturesof the bar cross section, as schematically shown in Fig. 11.Using computer-aided design (CAD) software to importphotosandthenamplifyingtentimesforprecisesizemeasure-ment, the grid of bending bar section is shown in Fig. 12.Taking the hash grid near pressure head the earnest forstrain analysis. In this study, the plane strain is assumed, and0.000.020.040.060.080.100.120.140.160100200300400500600700800Stress / MPaStrain 40Cr5etet0.000.020.040.060.080.10020040060080010001200Stress / MPastrain 42CrMo(a)(b)Fig. 9 Relationship curve of stress-strain: a 40Cr and (b) 42CrMoTable 2Mechanical properties of different material parameterss(MPa)E (GPa)B (MPa)n40Cr4102060.311820.2242CrMo9302100.31119.90.058051015202530051015202530 40Cr 42CrMoLoad /KNDisplacement /mmFig. 10 The load-displacement curveFig. 11 Diagram of taking photosInt J Adv Manuf Technol (2015) 79:151915291525bar bending deformation is mainly longitudinal deformationof fiber. Therefore, only the longitudinal strain state of grid isanalyzed. As shown in Fig. 12, choosing three column of thegrids, A, B, and C, to paint in CAD; comparing grid shapesbefore and back deformation; measuring the length of longi-tudinal grid lines of the three columns of the grids before andafter bending; and taking the natural logarithm of the gridlines length ratio after bending and before bending, the valueas the bar longitudinal strain is worked out. Setting the barfiber tensile strain as positive, and pressure strain as negative19.Thethree-columngridstraindataoftwobarsisshowninTable 3.(a)(b)Fig. 12 Grids of bar bending: a40Cr and b 42CrMoTable 3Strain data of bar grid linesRW=912.6112345678940CrA0.1370.1220.1060.0900.0740.0580.0420.0250.009B0.1250.1090.0940.0800.0650.0490.0350.0190.004C0.1100.0980.0850.0720.0580.0450.0330.0200.007101112131415161718A0.01050.02710.04460.06210.07940.09700.11360.13080.1471B0.01390.02910.04690.06120.07690.09190.10690.12270.1370C0.01260.02620.04100.05540.06990.08390.09780.11150.1251RW=800.9412345678942CrMoA0.0810.0710.0590.0480.0390.0290.0180.0070.0059B0.0950.0820.0700.0570.0430.0310.0170.0060.0064C0.0730.0660.0540.0440.0340.0230.0140.0040.00481011121314151617A0.01620.02660.03760.04780.05830.06810.07870.0900B0.01810.03050.04220.05450.06630.07810.09080.1041C0.01510.02520.03540.04630.05620.06550.07580.08611526Int J Adv Manuf Technol (2015) 79:15191529According to the location distribution of the grid lines lo-cated in the direction of bar diameter, from lower edge toupperedgeofthebar,thepositionstraincurveonthedirectionof bar diameter is shown in Fig. 13.Figure 13a is the strain distribution of 40Cr after bendingon the direction of radius. The coordinates of original neutrallayer is zero (before loading, the geometric neutral layer andthestrainneutrallayerarecoincident).Afterbending,itcanbeseen from the local amplification figure near the neutral layerthat the positive strain is generated on the position of originalstrainneutral layer. The strainvalue ofloweredgeofthe bar isnegative according to the deformation principle of continuity.There should be a fiber layer with a strain of 0 between thepositionoforiginalneutrallayer (0mm)andtheloweredgeofthe bar (14 mm), and this fibrous layer is strain neutral layerposition after bending. It is known that during the process of-15-10-5051015-0.15-0.10-0.050.000.050.100.15Line