

The entire automobile assembly is a key part of the automobile manufacturing process. The threaded connection is the most widely used connection process in the automobile industry. Whether the pre-tightening force of the thread pair connection is reliable, whether the assembly quality and operation of the tightening torque can be effectively guaranteed is mainly affected by two aspects: reasonable setting of the tightening torque design value of the assembly, and the tightening of the thread pair torque Effective control.
The analysis of the tightening torque process capability of the threaded pair connection is an assessment of the connection quality and the operational reliability guarantee ability of the automobile threaded pair. The two tie bolts used to fasten the urea tank in the automotive urea injection system discussed in this article are a common form of threaded connection in the industry.



Its structure is a T-bolt at each end and a middle bolt. The tension band (strip material) is welded and connected (as shown in Figure 1), but the current tightening torque of this structure lacks a complete analysis and calculation method, mainly referring to empirical formulas and process tests, and at the same time when tightening by the torque control method , Fastening characteristics are different from conventional fasteners. This increases the difficulty of clarifying the design value of the tightening torque and ensuring the reliability and quality of the assembly
Normality test
In the production process of automobile batch products with two-way tolerance, without special deviation, it generally obeys the normal distribution law, and its distribution curve is a Gaussian curve, denoted by X~N(μ,σ2). Due to the random sampling data of a large sample of n≥30, the process standard deviation (σ) in the steady state of the product is unknown.
We assume that the tightening torque of the tie bolt is measured randomly, which obeys the expected tightening value of the product (5Nm ) Requires a normal distribution X ~ N (μ, σ2), and the degree of confidence in the error is 95%; the statistics of the “t test” can be used for hypothesis testing (ie, the test of μ). The hypothesis test method is: replace the σ value of the product in the stable process with the standard deviation S of the sample, and make an unbiased estimate; it is concluded that the normality hypothesis is rejected