关于《燕尾形零件的精密测量》一文在本刊88年第5期上发表后,收到一些读者的来信,对此文的结论发表了一些异议。经仔细研究原文,认为文章确实存在问题。文中的Δα值只有当L或l中有一个数值为图上名义值时,才能满足l或L在公差范围内,故这种燕尾导轨的注法是不科学的。在实际使用过程中,燕尾导轨只接触3个面,因此,只需注l或L中一个尺寸的公差及给出角度公差,故另一端的尺寸是无关紧要的。而给出(L)及α的公差后,用一定直径D的滚柱进行测量时的测量值M的公差ΔM是较易确 After the article “Precise Measurement of Dovetail Parts” was published in Issue 5, Issue 88 in 1988, I received some letters from readers and made some objections to the conclusion of this article. After careful study of the original, that the article does exist problems. The value of Δα in this paper can only satisfy l or L within the tolerance range if there is a value in L or l for the nominal value on the graph. Therefore, the injection method of such a dovetail rail is unscientific. In actual use, the dovetail rails only touch 3 surfaces, so the tolerance of one dimension in L or L and the angular tolerance are given, so the size of the other end does not matter. Given the tolerances of (L) and α, the tolerance ΔM of the measured value M when measuring with a roller of a certain diameter D is more accurate