Tightening Factors and Dispersion in VDI 2230: Why Design Preload Must Be Discounted
Executive Summary
You spend ages on the calculation, yet the actual tightened preload never matches? The problem often lies in the tightening factor and preload dispersion. This article explains why design preload must be "discounted," and how to quantify and control dispersion under different tightening strategies.
Table of Contents
The Gap Between Design Value and Actual Value
The joint calculation gives a target preload, but the actual preload tightened on the shop floor is always a distribution, not a fixed value. The width of this distribution is the preload dispersion. It determines how much discount must be applied to the target preload during design, so that the joint remains safe in the worst case.
Understanding the sources of dispersion is the prerequisite for controlling it.
Four Sources of Dispersion
Source 1: Friction Coefficient Fluctuation
This is the dominant source. Fluctuation in the thread-pair friction coefficient and the bearing-surface friction coefficient is directly amplified into preload fluctuation. Under pure torque control, the relative preload fluctuation is approximately proportional to the relative friction-coefficient fluctuation.
Causes of friction-coefficient fluctuation include: inconsistent surface roughness, uneven coating thickness, differences in lubricant application amount, lubricant migration over storage time, cleanliness differences, and ambient temperature and humidity.
Source 2: Torque Application Accuracy
The tightening tool's torque accuracy, tightening speed, whether tightening is continuous, and whether it is staged all affect the final preload. The accuracy difference between a hand-held tool and a servoelectric spindle can be several-fold.
Source 3: Thread and Bearing-Surface Geometric Errors
Thread lead-angle deviation, thread profile error, face perpendicularity deviation, and bearing-surface flatness error all change the friction path and contact state.
Source 4: Material and Heat-Treatment Variation
Batch variation in bolt material yield strength and uneven hardness change the actual stress level at the same preload.
Tightening Factor: Converting Dispersion into Design Margin
VDI 2230 uses the tightening factor αA to quantify this effect. Its physical meaning is: the ratio of maximum assembly preload to minimum assembly preload.
αA = FM,max / FM,min
The relationships used in design are:
FM,min = νmin × F0.2 (minimum guaranteed preload)
FM,max = αA × FM,min ≤ 0.9 × F0.2 (maximum allowable preload)
In other words, the larger the tightening factor, the narrower the preload range that can be used safely. When αA is too large, a dilemma arises: "minimum preload is insufficient, but maximum preload already exceeds yield." The only way out is to improve the tightening process to shrink αA.
| Tightening Strategy | Typical αA | Usable Preload Range |
|---|---|---|
| Torque control, dry, hand | 1.5～2.0 | Very narrow |
| Torque control, lubricated + power tool | 1.3～1.5 | Medium |
| Torque control, lubricated + servoelectric | 1.2～1.4 | Medium-wide |
| Torque-angle method | 1.1～1.25 | Wide |
| Yield-point control | 1.05～1.15 | Widest |
Why Torque Control Has Such Large Dispersion
The root cause is that in the torque-preload relationship T = K × FM × d, K is a multiplicative factor. The fluctuation in K is amplified directly onto the preload.
For example: if K ranges from 0.18 to 0.24 (±14%), at a fixed torque the preload fluctuation range is the corresponding ±14% reverse fluctuation — a 1000 N target may land between 875 and 1200 N. In the extreme case where K is 0.16 to 0.28 (±27%), the preload fluctuation can reach ±27%.
The torque-angle method has small dispersion because it bypasses the friction coefficient: a small "seating torque" is first applied to eliminate clearance, then rotation continues by a fixed angle. The bolt elongation is precisely determined by the thread pitch and angle, and is essentially independent of the friction coefficient. This is why it is more accurate.
Four Paths to Reduce Dispersion
Path 1: Unify the Friction State
This is the highest cost-performance path. Specific measures include:
- Use fasteners with a lubricating coating applied uniformly at the supplier's factory, avoiding arbitrary on-site oil application;
- Clearly define and lock down the coating process parameters (film thickness, cure temperature, time);
- Control storage conditions and shelf life to prevent lubricant migration and loss of effectiveness.
Unifying the friction state typically reduces αA from 1.6～2.0 to 1.3～1.5 — a significant effect.
Path 2: Upgrade the Tightening Strategy
Upgrading from torque control to the torque-angle method or yield-point control can further reduce αA to 1.1～1.25. For high-safety-class joints (brakes, steering, battery pack fixation, load-bearing lifting points), this is a necessary investment.
Path 3: Improve Tightening Tool Accuracy
Use servoelectric tightening spindles to replace pneumatic or hand tools, and calibrate tools regularly. Also define the tightening speed — too-fast tightening causes dynamic effects that make the actual preload too high.
Path 4: Control Incoming-Material Consistency
Perform batch sampling of the bolt friction coefficient (i.e. torque-clamp force testing) to keep K within a narrow range. This test directly exposes incoming-material variation and is an effective means of mass-assembly quality control.
How αA Interacts with Design Margin
The tightening factor directly affects the usable design space of the joint. A simplified example illustrates this:
| Parameter | Option A (Dry, Hand) | Option B (Lubricated + Torque-Angle) |
|---|---|---|
| Yield-strength-based preload F0.2 | 100 kN | 100 kN |
| Minimum utilization factor νmin | 0.55 | 0.75 |
| Tightening factor αA | 1.8 | 1.15 |
| Maximum allowable preload (0.9 F0.2) | 90 kN | 90 kN |
| FM,min | 55 kN | 75 kN |
| FM,max = αA × FM,min | 99 kN ✗ exceeds limit | 86 kN ✓ compliant |
| Conclusion | Must lower νmin; usable preload is lower | Usable preload is 36% higher |
This example shows: with the same material, just by improving the tightening process, the usable preload can rise by more than one third. This is equivalent to improving the joint's anti-slip, anti-separation, and anti-fatigue capability without added cost.
Conclusion
The tightening factor and preload dispersion are the "invisible hands" in joint design. They do not appear on the part drawing, yet they determine whether the design value can be truly reproduced on the production line. The priority in engineering practice should be: first unify the friction state, then upgrade the tightening strategy, and only last consider enlarging the bolt size.
