Research On Assembly Precision Of CNC Machine Tools Considering Characteristic Parameters Of Joint

Some complex mechanical products require very high accuracy in assembly, suchas CNC machine tools. CNC machine tools may not successfully meet every precisiontarget in different assembly sequence. Some products in the failed assembly sequencewill need another manufacturing process to make sure their design accuracy. However,these cost lots of time and money.In order to assembly successfully at the first time, this paper analyzes three typesof error sources which have influence on the final assembly precision. When two partsare assembled in actual assembly, we can get sensitivity direction from assembly featureshape and fitting manner. Therefore, we can research the error caused by tolerancesband and clearances and the assembly sensitivity caused by tolerances band.This paper uses the finite element analysis software ABAQUS to solve the jointmaximum deformation so that we can research the joint deformation under differentbasic characteristic parameters. And a set of physical experimental data verify theaccuracy of this finite element model. Then we can get lots of joint deformation underdifferent sizes and quantity of basic characteristic parameters. For the purpose ofpredicting the joint deformation that widely existed in CNC machine tool assemblyprocess, this paper constructs BP neural network to forecast the effect of different basiccharacteristic parameters. Because of the limitations of the neural network, this paperchanges the hidden layer nodes and utilizes genetic algorithms to preprocess the initialpoint in the neural network. At last, the error between the output value and the actualvalue achieve the required precision.This paper proposes a new evaluation system about assembly precision afterdescribing each kind of error sources. First of all, the error sources must be projectedonto each assembly error propagation matrix through assembly coordinate systemconversion. Then the initial assembly point which we defined will pass assembly errorsto the required assembly feature or assembly features from different assemblysequences. According to coordinate differences between the required assembly features, we can calculate the values of closed cycle assembly precision and key assemblyprecision. The design weight must choose from the required design accuracy and thevalue of total assembly precision can be calculated. Finally the feasible region and theoptimal sequence of sequence groups can be concluded through comparing the threetypes of assembly precision. At last, this evaluation system about assembly precision isverified by an assembly example with two given assembly sequence.