Sign in

What’s new in Axomek 0.1.13

Released

Changes in detail

Results that change when a saved project is solved again:

  • Bolted joints with eccentric and a working moment take the opening limits at both edges of the interface (VDI 2230-1 Eq. 104, 105, 107 at U and at V): a negative workingMoment, or a positive one in case II of Table 4, opens the far edge V. Such joints need a larger clamp load FKA (before, it came out as zero), so FMmin, FMmax and the preload check rise, and FAab falls. openingMoment is now MBab in the sense of MB,max (it was always the positive limit, or infinite) and reverseOpeningMoment gives the other bound.
  • Eccentric joints and joints under a working moment get a new check noOpening: the clamp load at the least preload (Eq. 102) against the clamp load at which an edge opens (Eq. 107), under FA,max/MB,max and FA,min/MB,min. limits.noOpening false drops it (an accepted partial opening, Annex D).
  • Threads rolled after heat treatment take the mean load FSm = (FSAo + FSAu)/2 + FMzul (Eq. 172) with the working moment's share Φm*·MB/ssym and an opening interface's FSAKl: σASG changes for eccentric joints under a working moment or with a partially opening interface (Annex D).
  • The results gain amplitude (σSAo/σSAu, the axial and bending parts of σa and the working moment's share, FSm) and opening. Concentric joints without a moment are unchanged.
  • Axial thermal growth (settings.thermalGrowth) on a cross-located shaft: a bearing that holds the shaft one way only (each bearing of an X or O pair with arrangement set, a single angular contact or tapered roller bearing, a thrust bearing, an NJ or NF) now lifts off where the linear spring used to pull on the shaft: it carries no axial load, the other bearing takes it all, and the support reports an axialGap with a warning. With arrangement and preload set, the bearings' axial loads now include the mounted preload (before, only its thermal change), so their lives fall accordingly; an O pair whose growth exceeds the preload stops at zero preload in operation instead of going negative. Projects without thermal growth, and bearings that hold both ways (deep groove, four-point, double-row, spherical, NUP, matched pairs), are unchanged.
  • With a duty cycle, each load case's speed is checked against the thermally permissible speed nθ under that case's loads (duty.cases[] thermalPermissibleSpeed, thermalSpeedUtilisation): the check bearings.<id>.thermalSpeed comes back over the duty cycle as the worst case (before, the duty cycle's checks left it out), a warning only unless limits.permissibleTemperature is set. The bearing's own thermalSpeeds at the operating point are unchanged.
  • Ball screws with a geometry and a preload: the ball model's ratingFromGeometry is that of one nut of the double nut (the half with its share of the circuits), as C_a rates each nut and as the ISO 3408-5 ratings from the geometry count one nut's turns; before, it rated all the circuits as one nut, about 2^0.7 = 1.6 × too high, so calibratedLifeHours came out about 4 × too short (the linear-axis template: 29 216 h, now 121 164 h). Nuts without preload are unchanged.
  • Cylindrical gears rated by the 2019 edition with a ferritic nodular iron (GGG-ferritic) gear meshing with a surface-hardened one, and hardnesses given: the iron gear now takes the work hardening factor of ISO 6336-2:2019, 13.3.3 (ZW = 1.2 − (1.87·HB − 303.6)/1700, 1.0–1.2), as the pearlitic grade did; before, its ZW was 1. Its σHP and pitting safety rise by up to 20 %.
  • Cylindrical gears with details.limitedPitting and a gear between 6·10⁵ and 10⁹ contact cycles: σHP runs from its reference to its static value along both segments of the life curve (ISO 6336-2:2019, Formulae 7–10), as ZNT does; before, the share of the static factors ZW,stat against ZL·Zv·ZR·ZW ran in one straight line from 6·10⁵ to 10⁹ cycles. σHP changes the more, the further ZL·Zv·ZR·ZW lies from ZW,stat; it falls where ZL·Zv·ZR is below 1 (for example 596.8 → 589.0 MPa, −1.3 %, at 10⁸ cycles with ZL·Zv·ZR = 0.66). Bending, and contact without limited pitting, have one segment and are unchanged.
  • Cylindrical gears (and the planetary stages and gearboxes rated with them) without pitchDeviation, profileDeviation or helixDeviation take the tolerances of their accuracyGrade as ISO 1328-1:1995 tabulates them: its formulae at the geometric means of the ranges of module, diameter and face width, rounded (as ISO/TR 6336-30 rates its examples, and as fHβ5 of the face load factor already was); before, the formulae at the gears' own module, diameters and face width, unrounded. fpt, ffα and fHβ move by up to about ±10 %, and with them Kv, KHα, KHβ and the safeties: the templates' gear pair SH1 1.376 → 1.377 and SF1 4.200 → 4.205, the single-stage gearbox SH1 1.396 → 1.408 and SF1 4.401 → 4.475, the planetary stage's least SH 1.715 → 1.741 and SF 4.022 → 4.256. Deviations entered are unchanged.
  • Shafts: a load, support or spring closer than a twentieth of an element (1/3000 of the shaft's length with the default mesh) to a step, an end or a support now shares that node; a support or load moves onto it by at most that distance. Before, it made one very short element, and with shearDeformation off the reactions could come out wrong (a third too large, out of equilibrium, 0.001 mm from a step) or the shaft was refused as not held in place. Shafts whose points are further apart are unchanged.
  • Ball bearings without clearance (radialClearance 0) and an even number of balls, with conformities adding to 1.05 (0.52 and 0.53, as the presets), solve; before, they failed with "the equilibrium solver produced non-finite values".
  • Gear pairs, planetary stages and bevel pairs with viscosity40 at zero or below are refused with an error; before, they were rated with a lubricant factor ZL (and safeties) from no oil.
  • The subsurface stress map of a ball contact (the bearing's stress chart): its surface row now carries the surface stresses of the Hertz field (Johnson Eq. 3.41, 3.42 for a circle: σx = σy = −0.8·p0 against σz = −p0 at the centre, radial tension (1 − 2ν)/3·p0 at the edge and outside). Before, it took σx = σy = σz = −p (no shear, no von Mises stress, nothing outside the contact). The rows below the surface, the peaks and their depths, and roller (line) contacts are unchanged.