I use a graphical procedure implemented in CAXA to determine the measuring ball distance M for helical gears. The procedure is intended for involute helical gears whose transverse tooth profile can be generated by the involute gear module of the CAD software. I begin with the normal and transverse parameters of the helical gear, convert them into the corresponding transverse-plane values, and then construct the transverse tooth profile. Two measuring balls are represented in opposite tooth spaces by their actual circles and by imaginary circles whose diameter is adjusted for the base helix angle. The final dimension over the two balls is obtained directly from the drawing. My experience shows that this graphical method avoids involute function tables and iterative calculations, while giving results that are very close to those from analytical formulas.
Measuring tooth thickness by pins or balls is a common practice for cylindrical gears. In helical gears, span measurement may become impossible when the face width is smaller than W sin beta_b, where W is the span length and beta_b is the base helix angle. In such cases, the measuring ball distance M is an attractive alternative. The analytical calculation of M for helical gears usually requires involute function values or an iterative solution. Those steps are time-consuming and can introduce arithmetic or table-reading errors. I therefore reduce the problem to a precise transverse-plane construction in CAXA, where the involute gear module generates the correct transverse involute profile.
The most important idea in my method is that the physical measuring ball does not contact the helical tooth flank in the same way as a spur gear ball. The contact point lies outside the transverse plane that passes through the ball center. To represent the ball correctly in the transverse plane, I replace the actual ball by an imaginary circle whose diameter is larger than the actual ball diameter. This imaginary circle is tangent to the transverse involute profile. Once the imaginary circle centers are located graphically, I draw the actual ball circles and measure the outer distance between them. That distance is the required measuring ball distance M for the helical gears.
Nomenclature
| Symbol | Meaning | Unit |
|---|---|---|
| z | number of teeth | – |
| m_n | normal module | mm |
| m_t | transverse module | mm |
| alpha_n | normal pressure angle | degree or radian |
| alpha_t | transverse pressure angle | degree or radian |
| beta | helix angle at reference cylinder | degree or radian |
| beta_b | base helix angle | degree or radian |
| X_n | normal profile shift coefficient | – |
| d | reference diameter | mm |
| S_n | normal circular tooth thickness at reference cylinder | mm |
| S_t | transverse circular tooth thickness at reference cylinder | mm |
| S_t_bar | transverse chordal tooth thickness at reference cylinder | mm |
| d_p | actual measuring ball diameter | mm |
| d_pt | imaginary measuring circle diameter in transverse plane | mm |
| M | measuring ball distance | mm |
| R | radial distance from gear center to ball center in transverse plane | mm |
Fundamental Relations for Helical Gears
I start from the basic geometry of helical gears. The transverse module, transverse pressure angle, reference diameter, and transverse tooth thickness are obtained from the normal data by the following relations.
$$ m_t = \frac{m_n}{\cos \beta} $$
$$ \tan \alpha_t = \frac{\tan \alpha_n}{\cos \beta} $$
$$ \sin \beta_b = \sin \beta \cos \alpha_n $$
$$ d = m_t z = \frac{m_n z}{\cos \beta} $$
$$ S_n = m_n \left( \frac{\pi}{2} \pm 2 X_n \tan \alpha_n \right) $$
$$ S_t = \frac{S_n}{\cos \beta} = \frac{m_n \left( \frac{\pi}{2} \pm 2 X_n \tan \alpha_n \right)}{\cos \beta} $$
$$ \bar{S}_t = d \sin \left( \frac{S_t}{d} \right) $$
In the expression for S_n, the plus sign applies to an external helical gear and the minus sign applies to an internal helical gear. The transverse chordal tooth thickness S_t_bar is used in my graphical construction because it defines the correct chordal distance on the reference circle. This correction is important when the involute gear module of the CAD software does not reproduce the profile shift exactly in the transverse plane.
The imaginary measuring circle diameter is obtained from the actual ball diameter and the base helix angle:
$$ d_{pt} = \frac{d_p}{\cos \beta_b} $$
This equation is the key to reducing the three-dimensional ball contact of helical gears to a two-dimensional transverse construction. The imaginary circle is tangent to the transverse involute flanks at the same angular positions that correspond to the real ball contact. I then use the imaginary circle radius d_pt/2 as the offset distance from the corrected transverse flanks.
The base diameter and the involute function are also useful for understanding the construction:
$$ d_b = d \cos \alpha_t $$
$$ \text{inv} \alpha = \tan \alpha – \alpha $$
Here alpha is any pressure angle on the involute. The involute function appears in analytical solutions for the ball distance, but my graphical method does not require its numerical evaluation.
Normal and Transverse Parameter Conversion
| Normal or base parameter | Transverse parameter | Conversion relation |
|---|---|---|
| m_n | m_t | m_t = m_n / cos beta |
| alpha_n | alpha_t | tan alpha_t = tan alpha_n / cos beta |
| S_n | S_t | S_t = S_n / cos beta |
| beta | beta_b | sin beta_b = sin beta cos alpha_n |
| d_p | d_pt | d_pt = d_p / cos beta_b |
Why the Transverse Plane Is Sufficient for Helical Gears
The helical gear tooth surface is generated by a straight line on a generating plane that rolls on the base cylinder. The generating line is inclined to the base cylinder axis by the base helix angle beta_b. When the generating plane rolls, the line sweeps out an involute helicoid. The intersection of this helicoid with any transverse plane is an involute curve. Therefore, the transverse tooth profile of a helical gear is an involute. This is the geometric reason I can analyze the measuring ball in the transverse plane. The three-dimensional contact is converted into an equivalent two-dimensional contact through the imaginary circle diameter d_pt.
I do not need to draw the helical surface itself. I only need the transverse involute profile and the correct imaginary circle. This simplification is what makes the CAXA graphical method practical for helical gears. The helical nature of the gear is retained in the value of beta_b, which changes the imaginary circle diameter. All other construction steps are performed on the transverse profile.
Derivation of the Imaginary Circle Diameter
Consider the plane tangent to the base cylinder. The actual ball center is at some point outside the transverse plane of the contact point. The line from the ball center to the contact point is normal to the helical flank. This normal lies in the tangent plane and makes an angle beta_b with the transverse plane. The projection of the ball radius onto the transverse plane is larger than the actual radius. The right triangle formed by the actual radius, the projected radius, and the axial component gives
$$ \frac{d_{pt}}{2} = \frac{d_p}{2 \cos \beta_b} $$
and therefore
$$ d_{pt} = \frac{d_p}{\cos \beta_b} $$
This imaginary circle is the correct transverse representation of the actual measuring ball. If I mistakenly used d_p directly to offset the transverse flanks, the ball centers would be too close to the gear center, and the measured distance M would be too small. The error would be especially large for helical gears with a large helix angle.
Measuring Principle
I place two identical measuring balls in two opposite tooth spaces of the helical gear. For an even number of teeth, the two balls are placed in spaces separated by z/2 teeth. For an odd number of teeth, I use a span of (z-1)/2 teeth. In both cases, the balls are located near the middle of the tooth height, slightly toward the tip. The ball diameter can be selected according to common practice. For external helical gears, a useful starting value is d_p = k_p m_n with k_p between 1.65 and 2.00. For internal helical gears, a value near d_p = 1.65 m_n is often used. The final choice should ensure that the ball contacts the involute flank and not the root fillet or the tip corner.
The physical ball contacts the helical flank along a curve, and the tangent point does not lie in the transverse plane through the ball center. The normal plane tangent to the base cylinder is used to relate the actual ball to the transverse profile. In that plane, the ball center, the actual contact point, and the imaginary contact point form a right triangle. The projected imaginary circle in the transverse plane has diameter d_pt. This is why I never use the actual ball diameter directly to offset the transverse involute profile. Instead, I use d_pt/2 to find the ball center, and then I draw the actual ball with diameter d_p for the final distance measurement.
After the two ball centers O1 and O2 are located, I connect them and extend the line to the far intersections with the two ball circles. The distance between those far intersections is the measuring ball distance M. Mathematically, this is
$$ M = O_1 O_2 + d_p $$
For an even number of teeth, the two ball centers are diametrically opposite, so
$$ M = 2R + d_p $$
For an odd number of teeth with a span of (z-1)/2, the angular separation between the two ball centers is gamma = pi – pi/z. Therefore,
$$ O_1 O_2 = 2R \cos \left( \frac{\pi}{2z} \right) $$
and
$$ M = 2R \cos \left( \frac{\pi}{2z} \right) + d_p $$
Here R is the radial distance from the gear center to each ball center in the transverse plane. These equations provide a useful check on the graphical result, although the graphical method itself does not require them.
Basic Parameters Needed for the Graphical Method
Before I open CAXA, I compute the parameters that control the transverse involute profile and the imaginary circle. The following table summarizes the complete set.
| Parameter | Purpose in the graphical method | Formula or source |
|---|---|---|
| m_t | transverse module used by the gear module | m_t = m_n / cos beta |
| alpha_t | transverse pressure angle for the involute profile | tan alpha_t = tan alpha_n / cos beta |
| z | number of teeth and span selection | given |
| X_n | profile shift for tooth thickness | given |
| d | reference diameter for chordal correction | d = m_t z |
| S_t | transverse circular tooth thickness at reference circle | S_t = S_n / cos beta |
| S_t_bar | transverse chordal tooth thickness for correction | S_t_bar = d sin(S_t/d) |
| beta_b | base helix angle for imaginary circle | sin beta_b = sin beta cos alpha_n |
| d_pt | imaginary circle diameter for ball center location | d_pt = d_p / cos beta_b |
| d_p | actual ball diameter for final distance | selected or given |
Example Data and Computed Parameters
I use the following external helical gear as a worked example. The data are typical for a measurement problem involving helical gears.
| Given parameter | Value |
|---|---|
| z | 35 |
| m_n | 2.5 mm |
| alpha_n | 20 degrees |
| beta | 19 degrees |
| X_n | +0.112 |
| d_p | 4.7625 mm |
From these data, I compute the transverse and base parameters that are needed for the graphical construction.
| Computed parameter | Formula | Value |
|---|---|---|
| beta_b | arcsin(sin beta cos alpha_n) | 17.814 degrees |
| d_pt | d_p / cos beta_b | 5.00233879 mm |
| alpha_t | arctan(tan alpha_n / cos beta) | 21.053834 degrees |
| m_t | m_n / cos beta | 2.644051703 mm |
| d | m_t z | 92.5418096 mm |
| S_t | m_n(pi/2 + 2 X_n tan alpha_n) / cos beta | 4.368834473 mm |
| S_t_bar | d sin(S_t/d) | 4.3672 mm |
These values are used directly in the CAXA drawing. I keep all decimal places during calculation and enter the values into the software with as much precision as the dialog allows. The imaginary circle radius used for the offset is half of d_pt, namely 2.501169395 mm. The actual ball radius used for the final circles is half of d_p, namely 2.38125 mm.
Step-by-Step Graphical Procedure in CAXA
I now describe the exact sequence I use in CAXA. The sequence is written for external helical gears. For internal helical gears, the current involute gear module in some CAXA versions is not accurate enough, so I do not recommend this graphical procedure for internal helical gears until the software is improved.
- Open CAXA and start the gear module. I open the drawing environment and select the involute gear tool from the advanced drawing panel.
- Enter transverse parameters, not normal parameters. In the involute gear parameter dialog, I enter m_t, alpha_t, z, and the transverse profile shift. I use a temporary profile shift of zero because the software may not reproduce the exact transverse chordal tooth thickness. The chordal correction is performed later in the drawing.
- Set the tooth form options. I set the tip fillet radius to zero, keep the root fillet small, uncheck the effective tooth count option, set the graphical precision to 0.001 mm, and enable extended centerlines.
- Generate the transverse tooth profile. After confirming, CAXA draws the transverse involute profile of the helical gear. This profile is the starting point for the graphical measurement.
- Draw the reference circle. I draw a circle with diameter d equal to the transverse reference diameter. This circle is used to correct the chordal tooth thickness and to identify the measuring tooth spaces.
- Select the two measuring tooth spaces. For an odd tooth count, I use a span of (z-1)/2 teeth. For an even tooth count, I use a span of z/2 teeth. I mark the two opposite spaces as A and B. The balls will be placed in these spaces.
- Decompose the gear profile block. The generated gear profile is usually inserted as a block. I decompose it so that individual involute flanks can be offset and edited.
- Correct the transverse chordal tooth thickness. In space A, I find the intersection point C of the reference circle with the right flank of the right tooth that bounds the space, on the side toward the tooth tip. I draw a circle centered at C with radius equal to the transverse chordal tooth thickness S_t_bar. This circle intersects the reference circle at point G on the space side. I then use the offset-line command in through-point mode to offset the opposite involute flank D through point G, producing a new flank D prime. I delete the original flank D. I repeat the same correction for space B using the corresponding points E, H, and flank F prime. This step ensures that the transverse circular tooth thickness at the reference circle matches the calculated value.
- Locate the imaginary circle centers. In space A, I offset the two corrected involute flanks M and D prime into the space by the imaginary circle radius d_pt/2. The two offset lines intersect at O1. In space B, I offset the two corrected flanks N and F prime by the same distance. Their intersection is O2. O1 and O2 are the centers of the imaginary measuring circles in the transverse plane.
- Draw the actual measuring balls. I draw circles centered at O1 and O2 with diameter equal to the actual ball diameter d_p. These circles represent the real measuring balls in the transverse projection.
- Measure the ball distance M. I connect O1 and O2 with a line and extend it to the far intersections with the two ball circles. I place a dimension between those two far intersections. The dimension is the measuring ball distance M.
- Verify the result. I compare the graphical M with an analytical value if one is available. In my example, the graphical result is 100.5395 mm and the analytical result is 100.539 mm, a difference of only 0.0005 mm.
I have found that the chordal correction in step 8 is essential. Without it, the generated transverse profile may have a reference-circle tooth thickness that is slightly different from the calculated value. This difference can shift the ball centers and produce an error of several hundredths of a millimeter or more. The correction is especially important for profile-shifted helical gears, because the standard gear module may not apply the profile shift in the same way as the analytical transverse calculation.
CAXA Settings for the Example
| Setting | Value used | Reason |
|---|---|---|
| Gear type | External | Method validated for external helical gears |
| Module | 2.644051703 mm | Transverse module m_t |
| Pressure angle | 21.053834 degrees | Transverse pressure angle alpha_t |
| Number of teeth | 35 | Given |
| Profile shift | 0 | Correct later by chordal thickness |
| Tip fillet | 0 | Avoid interference with ball contact |
| Root fillet | Small | Keep root clearance |
| Effective teeth | Unchecked | Full profile display |
| Precision | 0.001 mm | Accurate graphical result |
| Centerline | Extended | Easy alignment and symmetric construction |
Graphical Result and Comparison
After completing the construction, I obtain the following result for the worked example.
| Method | M value | Difference from analytical | Relative error |
|---|---|---|---|
| Graphical CAXA method | 100.5395 mm | 0.0005 mm | 0.000497 percent |
| Analytical formula | 100.539 mm | reference | reference |
The graphical result is well within the precision required for normal gear measurement. The small difference comes from the finite precision of the CAD construction and from the rounding of the analytical value. I can further reduce it by setting a smaller graphical precision and by dimensioning the offset lines and circles with more decimal places.
Mathematical Verification of the Graphical Construction
I verify the graphical construction by deriving the relationship between the ball centers and the measured distance. Let O be the gear center. Let O1 and O2 be the two ball centers in the transverse plane. Let R be the radial distance from O to each ball center. Because the two measuring spaces are symmetric, R is the same for both balls. Let gamma be the angular separation between the two ball centers.
$$ O_1 O_2 = \sqrt{R^2 + R^2 – 2R^2 \cos \gamma} = 2R \sin \left( \frac{\gamma}{2} \right) $$
The measuring ball distance is the distance between the outer extremities of the two balls along the line O1O2:
$$ M = O_1 O_2 + d_p $$
For an even number of teeth, the two spaces are diametrically opposite, so gamma = pi and
$$ M = 2R + d_p $$
For an odd number of teeth, if the span is (z-1)/2, the angular separation is gamma = pi – pi/z. Hence,
$$ M = 2R \cos \left( \frac{\pi}{2z} \right) + d_p $$
This derivation confirms that the graphical construction measures the correct transverse distance. The only remaining requirement is that the ball center radius R be determined from the correct imaginary circle tangency. This is exactly what the offset by d_pt/2 accomplishes.
I also note that the base helix angle enters the construction only through the imaginary circle diameter. The transverse profile itself is generated with the transverse parameters. This separation of the three-dimensional ball geometry into a transverse imaginary circle is the central simplification of my method.
Span Selection for Odd and Even Tooth Counts
| Tooth count parity | Span in teeth | Angular separation gamma | O1O2 expression |
|---|---|---|---|
| Even | z/2 | pi | 2R |
| Odd | (z-1)/2 | pi – pi/z | 2R cos(pi/(2z)) |
Precision and Error Analysis
I have examined the main sources of error in the graphical method. The method is not a replacement for a certified metrology calculation, but it is more than accurate enough for ordinary design, process planning, and first-article checks. The following table lists the error sources I have observed and the ways I reduce them.
| Error source | Typical effect on M | Mitigation |
|---|---|---|
| Finite CAD precision | 0.0001 to 0.001 mm | Set precision to 0.001 mm or smaller |
| Rounding of input parameters | 0.0005 to 0.005 mm | Use full calculator precision before entering values |
| Incorrect chordal correction | 0.01 to 0.10 mm | Use S_t_bar and through-point offset |
| Block not decomposed | prevents accurate editing | Decompose the gear profile block |
| Offset on wrong side | large error | Offset into the tooth space |
| Wrong ball diameter | proportional error | Verify d_p and d_pt |
| Wrong span for odd teeth | shift in ball centers | Use (z-1)/2 for odd z |
| Base helix angle error | error in d_pt | Calculate beta_b carefully |
For the worked example, the relative difference between the graphical and analytical results is about 0.0005 percent. This level of agreement is typical when the drawing is prepared carefully. I can improve the agreement by using a precision of 0.0001 mm, by dimensioning the construction with more decimal places, and by avoiding unnecessary rounding of intermediate values.
Comparison with Analytical Methods
Traditional analytical methods for the measuring ball distance of helical gears require the evaluation of involute functions. For example, the pressure angle at the ball center is usually found from an equation of the form
$$ \text{inv} \alpha_M = \text{inv} \alpha_t + \frac{d_{pt}}{d \cos \alpha_t} – \frac{S_t}{d} + \frac{\pi}{z} $$
or a similar relation, depending on the sign convention and whether the gear is external or internal. The exact form varies with the measurement arrangement and the definition of the tooth space. In practice, I have to solve such equations by table lookup or iteration. That process is slower and more error-prone than the graphical construction. My graphical method avoids the involute function table entirely. It also avoids the need to distinguish between standard and profile-shifted external helical gears, and it handles both odd and even tooth counts with the same drawing procedure.
I can summarize the comparison in the following table.
| Aspect | Analytical method | My graphical method |
|---|---|---|
| Involute function table | Often required | Not required |
| Iteration | Often required | Not required |
| Visual check | Difficult | Natural |
| Odd/even tooth count | Separate formulas | Same procedure with different span |
| Standard/profile-shifted external gear | Separate formulas | Same procedure with chordal correction |
| Internal helical gear | Possible with correct formulas | Not recommended in CAXA 2009 due to module limitations |
| Typical error | Very small | Very small, about 0.0005 percent in the example |
| Speed | Moderate | Fast after the first setup |
Advantages of the Graphical Method
I use the graphical method for helical gears because it has several practical advantages.
- It gives a clear visual picture of the ball position in the tooth space.
- It eliminates the need for involute function tables and iterative solutions.
- It reduces the chance of sign errors in the analytical formulas.
- It allows me to check the construction directly on the drawing.
- It handles odd and even tooth counts with a simple span rule.
- It can be applied to standard and profile-shifted external helical gears after the chordal correction.
- It produces results that are very close to analytical values.
- It is fast once the basic parameters are computed.
Limitations and Cautions
I also recognize the limitations of the method.
- The current involute gear module in some CAXA versions is not accurate enough for internal helical gears. I therefore do not recommend the graphical method for internal helical gears until the software is improved.
- The method depends on correct input parameters. If m_n, alpha_n, beta, X_n, or d_p is wrong, the graphical result will be wrong.
- The chordal correction is essential. Skipping it can produce a noticeable error.
- The drawing precision must be set high enough. A coarse precision can hide small but important offsets.
- The method is not a substitute for a certified analytical calculation when extreme accuracy is required.
- The ball diameter must be chosen so that contact occurs on the involute flank. If the ball is too large or too small, the contact may move to the tip or root.
Practical Recommendations
On the basis of my experience, I recommend the following practices for measuring ball distance of helical gears by the CAXA graphical method.
- Compute all transverse parameters before starting the drawing. Use full precision in the calculations.
- Enter transverse module and transverse pressure angle into the gear module, not normal values.
- Use a temporary profile shift of zero and correct the chordal tooth thickness later.
- Set the tip fillet to zero and keep the root fillet small.
- Uncheck effective tooth count and set the graphical precision to 0.001 mm or better.
- Draw the reference circle with the exact transverse reference diameter.
- Choose the correct span: z/2 for even teeth and (z-1)/2 for odd teeth.
- Decompose the gear profile block before offsetting flanks.
- Use the transverse chordal tooth thickness S_t_bar to correct the measuring tooth spaces.
- Offset the corrected flanks by d_pt/2 to find the ball centers.
- Draw the actual ball circles with diameter d_p and dimension the far intersections.
- Compare the graphical result with an analytical result for the first article or for a new gear design.
Common Mistakes and How I Avoid Them
| Mistake | Consequence | Prevention |
|---|---|---|
| Using normal module in the transverse gear module | Wrong transverse profile | Compute m_t = m_n / cos beta |
| Using normal pressure angle instead of transverse pressure angle | Wrong involute shape | Compute alpha_t from tan alpha_t = tan alpha_n / cos beta |
| Forgetting to decompose the gear block | Cannot offset individual flanks | Decompose the block before editing |
| Offsetting on the wrong side | Ball centers outside the space | Offset into the tooth space |
| Using d_p instead of d_pt for ball center location | M is too small | Use d_pt = d_p / cos beta_b |
| Forgetting chordal correction | Reference tooth thickness is wrong | Use S_t_bar and through-point offset |
| Using z/2 span for odd teeth | Wrong angular separation | Use (z-1)/2 for odd z |
| Setting precision too low | Graphical error increases | Set precision to 0.001 mm or better |
| Drawing actual ball before imaginary circle | Wrong ball center | Locate imaginary circle centers first |
| Not extending centerline | Difficult alignment | Enable extended centerline |
Quality Control Checklist
| Check | Requirement |
|---|---|
| Input parameters | m_t, alpha_t, z, beta_b, d_p, d_pt, S_t_bar verified |
| Gear module settings | Transverse values used, tip fillet zero, precision high |
| Reference circle | Diameter equals d exactly |
| Measuring spaces | Span selected according to odd/even tooth count |
| Chordal correction | Applied to both measuring spaces |
| Imaginary circle offset | d_pt/2 used, not d_p/2 |
| Ball circles | Diameter equals d_p, centered at O1 and O2 |
| Dimension | Measured between far intersections along O1O2 |
| Verification | Compared with analytical value when available |
Comparison with Other Tooth Thickness Measurement Methods
| Method | Condition | Advantage | Limitation |
|---|---|---|---|
| Span measurement W | Face width greater than W sin beta_b | Simple and direct | Not possible for narrow helical gears |
| Chordal tooth thickness | Reference circle accessible | Simple calculation | Less accurate for helical gears |
| Ball measurement M | Balls fit in tooth spaces | Works for narrow gears | Requires ball selection and calculation |
| Graphical ball measurement | CAD with involute gear module | No involute tables, visual, fast | Requires careful setup and chordal correction |
Application Notes for Different Helical Gear Types
My graphical method is intended primarily for external helical gears. I have used it for standard gears and for profile-shifted gears. The profile shift is handled through the transverse chordal tooth thickness correction. The method also works for both odd and even tooth counts. The only difference is the span used to select the two measuring spaces. For even tooth counts, the two spaces are diametrically opposite. For odd tooth counts, the two spaces are separated by (z-1)/2 teeth, which introduces a small angular correction that is automatically included in the graphical dimension.
For internal helical gears, the situation is different. The involute gear module in some CAXA versions does not generate the internal transverse profile with sufficient accuracy for this measurement. The root and tip geometry of internal gears also changes the ball contact condition. For these reasons, I do not use the graphical method for internal helical gears. I prefer analytical formulas for internal helical gears until the CAD module is improved and validated.
Implementation Notes for CAXA
When I implement the method in CAXA, I follow a fixed command sequence. I open the involute gear tool and enter the transverse parameters. In the tooth form preview dialog, I set the tip transition fillet radius to zero, reduce the root transition fillet radius, uncheck the effective tooth count option, set the precision to 0.001 mm, and enable the extended centerline. After the profile is generated, I draw the reference circle with the circle command. I decompose the gear profile block. I use the offset command in through-point mode for the chordal correction. I use the offset command with a fixed distance of d_pt/2 to locate the imaginary circle centers. I use the circle command to draw the actual balls. Finally, I use the dimension command to measure the distance between the far intersections of the two ball circles.
I also keep a written record of the computed values: m_t, alpha_t, beta_b, d_pt, d, S_t, S_t_bar, and the final M. This record makes it possible to reproduce the drawing and to check the result later. For helical gears with different helix angles or profile shifts, the same procedure is repeated with the new transverse parameters.
Mathematical Appendix: Involute and Base Circle Relations
For completeness, I include the basic involute relations that underlie the transverse profile. The base diameter is
$$ d_b = d \cos \alpha_t $$
The base radius is
$$ r_b = \frac{d_b}{2} $$
The involute function at any pressure angle alpha is
$$ \text{inv} \alpha = \tan \alpha – \alpha $$
At the reference circle, the transverse pressure angle is alpha_t. The transverse tooth thickness at the reference circle is S_t. The chordal tooth thickness S_t_bar is used because the graphical construction works with chord lengths on the reference circle. The chordal value is
$$ \bar{S}_t = d \sin \left( \frac{S_t}{d} \right) $$
These relations are independent of the measuring ball. They define the transverse involute profile that CAXA generates and corrects. The ball measurement then adds the imaginary circle and the actual ball circles.
Final Remarks
I have described a graphical method for obtaining the measuring ball distance M of helical gears in CAXA. The method begins with the basic normal and transverse parameters of the helical gear. It uses the involute gear module to generate the transverse tooth profile, corrects the chordal tooth thickness at the reference circle, locates the imaginary ball centers by offsetting the corrected flanks by d_pt/2, draws the actual ball circles, and dimensions the far intersections. The result is the measuring ball distance M.
The method avoids involute function tables and iterative calculations. It is clear, visual, fast, and less error-prone than the traditional analytical route. In the worked example, the graphical result was 100.5395 mm and the analytical result was 100.539 mm, a difference of only 0.0005 mm. For external helical gears, including standard and profile-shifted designs with odd or even tooth counts, the method is practical and accurate. For internal helical gears, I do not recommend the graphical method in CAXA 2009 because the current gear module is not accurate enough. With careful parameter calculation, high drawing precision, and the chordal correction, the graphical method provides a reliable way to obtain the measuring ball distance for helical gears.
In my daily work, this graphical approach has become a standard tool for checking the measuring ball distance of helical gears. It gives me confidence that the ball is placed correctly in the tooth space, that the transverse profile is correct, and that the final dimension is consistent with the analytical result. When I need to evaluate a new helical gear design, I compute the transverse parameters, generate the profile, apply the chordal correction, locate the imaginary circle centers, draw the actual balls, and read the distance M. The entire process is transparent and easy to review. This transparency is the main reason I prefer the graphical method for helical gears.