Struve Function H_v(x) Calculator

Struve Function H_v(x) Calculator

Emerging AI Skill PRO

Enter Calculation

Both inputs are dimensionless real numbers. For non-integer order, use x ≥ 0.

Formula

Hv(x)=∑k=0∞(−1)kΓ(k+32)Γ(k+v+32)(x2)2k+v+1\mathbf{H}_{\text{v}}\!\left(\text{x}\right) = \sum_{k=0}^{\infty} \frac{(-1)^{k}}{\Gamma\!\left(k+\frac{3}{2}\right)\,\Gamma\!\left(k+\text{v}+\frac{3}{2}\right)} \left(\frac{\text{x}}{2}\right)^{2k+\text{v}+1}
Usage stats

Advertisement

Results

Struve Function Hv(x)
0.568656627
dimensionless
Method Power-series definition (Lanczos gamma)
Note Accuracy reduces for large |x|; use x ≥ 0 for non-integer order.

Report an error on this page

What is the Struve function?

The Struve function, written \(\mathbf{H}_{v}(x)\), is a special function that appears throughout mathematical physics. It is named after the astronomer Hermann Struve and arises naturally as a particular solution of the inhomogeneous Bessel differential equation. You will meet it in acoustics (radiation from a vibrating piston), fluid dynamics, electromagnetics, and unsteady aerodynamics. This calculator evaluates \(\mathbf{H}_{v}(x)\) for any real order v and argument x.

Graph of Struve function curves for several orders showing decaying oscillation
The Struve function H_v(x) rises then oscillates with slowly decaying amplitude.

The formula

The Struve function is defined by the convergent power series

$$\mathbf{H}_{\text{v}}\!\left(\text{x}\right) = \left(\frac{\text{x}}{2}\right)^{\text{v}+1} \times \sum_{k=0}^{\infty} \frac{(-1)^{k}\left(\frac{\text{x}}{2}\right)^{2k}}{\Gamma\!\left(k+\frac{3}{2}\right)\cdot\Gamma\!\left(k+\text{v}+\frac{3}{2}\right)},$$

where \(\Gamma\) is the gamma function. The series converges for every real x, although for large |x| many terms are required and an asymptotic expansion is more efficient. This tool sums the series directly, using a Lanczos approximation for the gamma factors and a stable term-ratio recurrence to avoid overflow.

Advertisement
Diagram of an alternating power series with shrinking terms summing to a value
H_v(x) is evaluated as a convergent alternating power series in k.

How to use it

Enter the order v (any real number) and the argument x. For non-integer order, keep x \(\ge\) 0, because \(\left(\frac{x}{2}\right)^{v+1}\) becomes complex for negative x. Press calculate to get \(\mathbf{H}_{v}(x)\). Both inputs are pure dimensionless numbers.

Worked example

Take the default v = 0, x = 1. Here half = 0.5, half² = 0.25 and the prefactor \(\left(\frac{x}{2}\right)^{v+1} = 0.5\). The first series terms are \(1.273239545,\ -0.141471061,\ 0.005658842,\ -0.000115487,\ \ldots\) summing to about \(1.137313265\). Multiplying by the prefactor \(0.5\) gives $$\mathbf{H}_{0}(1) \approx 0.5686566,$$ matching the known high-precision value \(0.56865663339780\).

Advertisement

Interpreting the Result and Its Accuracy Limits

Small x. The leading term of the series dominates, so \(\mathbf{H}_v(x)\approx \dfrac{(x/2)^{v+1}}{\Gamma(3/2)\,\Gamma(v+3/2)}\). The function grows like a power law \(x^{v+1}\) near the origin. For v = 0 this gives \(\mathbf{H}_0(x)\approx 2x/\pi\), and for v = 1 it gives \(\mathbf{H}_1(x)\approx 2x^2/(3\pi)\).

Large x. For fixed v and \(x\to\infty\),

$$\mathbf{H}_v(x)\approx Y_v(x)+\frac{(x/2)^{v-1}}{\sqrt{\pi}\,\Gamma\!\left(v+\tfrac12\right)}.$$

So \(\mathbf{H}_v(x)\) follows the oscillating Bessel function \(Y_v(x)\) plus a slowly varying algebraic term. The difference \(\mathbf{H}_v(x)-Y_v(x)\) has its own asymptotic series in inverse powers of x. For v = 0 it is \(\tfrac{2}{\pi}\left(\tfrac1x-\tfrac1{x^3}+\tfrac{9}{x^5}-\cdots\right)\).

Cancellation error. The power series converges for every x, but its terms alternate in sign. Their magnitudes grow to roughly \(e^{x}\) in size before they decay, while the true value is only of order 1 (or smaller). In double precision (about 16 digits), roughly \(x/\ln 10\approx 0.43\,x\) digits are lost. Around |x| ≈ 20–30 most or all significant digits are gone. Beyond that range the series result should not be trusted. Use the asymptotic expansion of \(\mathbf{H}_v(x)-Y_v(x)\) together with an accurate \(Y_v(x)\), or use higher-precision arithmetic.

Order and domain. For non-integer v the factor \((x/2)^{2k+v+1}\) is real only for \(x\ge 0\) (principal branch). When \(v+\tfrac32\) is a non-positive integer, the gamma function in the denominator has poles and the corresponding terms vanish.

FAQ: Why does the calculator give unreliable values for large x?

The alternating series suffers catastrophic cancellation. Individual terms become far larger than the final sum, and floating-point rounding errors in those large terms swamp the small true result. This is a limit of the numerical method, not of the function itself. The remedy is an asymptotic expansion, the relation to \(Y_v\), or extended-precision arithmetic.

FAQ

Can v be negative or fractional? Yes. The series is valid for any real v, except where \(v + \frac{3}{2}\) is a non-positive integer (\(v = -1.5,\ -2.5,\ \ldots\)); there the reciprocal gamma simply makes those terms vanish, which is handled automatically.

What happens at x = 0? For \(v > -1\) the function is 0 because the \(\left(\frac{x}{2}\right)^{v+1}\) prefactor vanishes; the calculator returns 0.

Is it accurate for large x? The power series loses precision through cancellation for |x| beyond roughly 30. For very large arguments an asymptotic expansion gives better results.

Last updated:

Most popular in Math and Statistics

View all Math and Statistics calculators