;Compared with spur gears, helical gears have the advantages of relatively higher contact ratio and load capacity, as well as more smooth and silent operation. As a consequence, helical gears ar the most used elements for power transmission.
In general, the shafts are deformed caused by the loads acting in the gears. In order to explore the effects of the gear shaft deformation on the contact characteristics of loaded teeth, an LTCA (loaded tooth contact analysis) model are developed in the thessis. The proposed LTCA model, based on the influence coefficient method, involves the contacted deformation and the bending of the tooth, as well as the torsion and deflection of the shaft. This LTCA model includes also a tooth contact model for gear meshing analysis, in which the mathematical equations of the actual helical gear flank surfaces, the assembly conditions of the gear pair, the contact points and the tooth gaps can be involved. Because the actual three dimensional flank surfaces are considered in the LTCA model, the contact stress distribution and the also the corresponding contact patterns can be simulated, even also the non-hertz contact of the the engaged teeth.
Three different tooth forms of the helical gears are considered in the study: non-modified flanks, ideally double corwned flanks and double crownded flanks by profile-grinding. On the other hands, two support arrangements of the shafts are also involved in the analysis: simple support arrangement and overhand arrangement. The contact characteristics, analyzed in the thessis, include the contact stress distribution, the variation of the load sharing and the max. contact stress during the gear meshing, as well as the transmission errors.
It can be recognized from the analysis results of the non-modified helical gears that the contact stress near the face-end of the bearing-side increases obviously due to the effect of shaft deformation. This phenomenon of uneven stress distribution is more serious with the increasing of shaft deformation (inclination angle). On the other hand, the concentrated stresses can be also found in the tooth conact on the tip and the face-end. The peak-to-peak value and the average amount of the loaded transmission error in this case increases also by the enlarged deformation of the loaded shaft.
In order to increase the load capacity and elimate the stress concentration, the flank modification of helical gears is, an essentail but alos effective method. In the study, double crowning modification is choosen, which consists of the helix modification, the lead crowning modification, and the profile crowning crowning modification. Helix modification can reduce the uneven contact stress distribution on flanks due to the shaft doformation, while lead and profile crowning can elimate stress concentration on the face-end, and tip/root of flanks. The analysis results of the ideally modified helical gears show that an uniform oval-shaped contact pattern is performed. The stress concentration on the flanks during the gear meshing does not occur, but the loaded contact ratio decreases from 2.47 (line contact) to 1.76 (point contact). In addition, the variation of the load sharing of the single tooth pair during the gear meshing becomes smooth.
Profile grinding is now a widely applied finishing method for helical gears, because of its high precision and good productivity. However, twisted tooth flank occurs if there are no correction measures conducted for profile grinding of the lead crowned flnaks. The contact characteristics of the double crowned flanks with nature twist are thus studied by using the proposed LTCA approach. Because the ease-off twisted right flank of gears with right hand helix angle is similar to the diagonal flank modification, the contact characteristics of this flank side are better than those of the left flank. Similarly, this guidline is also valid for the gear with left hand helix angle, where the contact characteristics of the left flank is better than those of the right flank. Besides, the peak-to-paek value of twisted flanks with the modification parameters used in the case is twice larger than that of ideally modified flank, but the averge value is smaller.