Yamagishi Complexity Torus LabYamagishi Lab

[Paper 637] For a Circular Orbit More Than 11.94 Times Farther Out, Going Even Farther First and Coming Back Needs Less Fuel

Abstract

A craft moves from a circular orbit of radius $1$ to one of radius $R$. The Hohmann transfer goes in two burns, one onto an ellipse and one to circularise. The bi-elliptic transfer goes in three: onto an ellipse with apoapsis $B$ ($B>R$), a burn at $B$ raising the periapsis to $R$, and a burn at $R$ to circularise. Since more fuel is needed the larger the total change of velocity $\Delta v$, the transfers are compared by total $\Delta v$ (with the speed on the starting circle as $1$). No new theorem or law is claimed.

First, there are two boundaries——the bi-elliptic transfer with $B$ taken ever farther equals the Hohmann transfer at $R=\mathbf{11.938765}$. The Hohmann $\Delta v$ reaches its maximum of $0.536258$ at $R=15.581719$, and for targets beyond that the bi-elliptic transfer was cheaper for every $B$. At $R=13$, between the two boundaries, it is cheaper only when $B$ lies beyond $48.904843$ (Sections 2 and 3).

Second, and this is the core. The Hohmann $\Delta v$ does not keep rising as the target gets farther——beyond $R=3.304167$, the $\Delta v$ to enter the target circular orbit exceeds the $\Delta v$ to escape to infinity from the start ($\sqrt{2}-1=0.414214$). Entering a far circular orbit costs more than escaping outright. So a path going out almost far enough to escape, then coming back, can be cheaper (Section 2).

Third, set side by side, the bi-elliptic transfer burns more at the start and a little far out——at $R=20$ and $B=100$, the Hohmann transfer takes $0.380131+0.154600=0.534731$ and the bi-elliptic $0.407195+0.043663+0.065068=0.515927$, $0.018805$ less. But it takes $24.207343$ times as long (Section 4).

Fourth, the formula’s $\Delta v$ really arrives——integrating the orbits numerically, both transfers, with the formula’s $\Delta v$, passed through their apoapses and periapses and reached the circular orbit of radius $20$ within $10^{-5}$. From a $300$ km circular Earth orbit to a circular orbit at the Moon’s distance ($R=57.560963$), the Hohmann transfer needs $3.936544$ km/s against $3.621910$ km/s for the bi-elliptic transfer going far out (Sections 5 and 6).

Keywords: Hohmann transfer, bi-elliptic transfer, orbital transfer, delta-v, orbital mechanics.

Scope (what is not claimed)

No new theorem or law is claimed——the Hohmann transfer (Hohmann, $1925$), the bi-elliptic transfer (Sternfeld, $1934$), and the two boundaries $11.94$ and $15.58$ are known. Real launch planning is not discussed——time constraints, the gravity of other bodies, and the duration of burns are not included; for time only the ratio is given. Why the bi-elliptic transfer gains is not pinned to a single reason——only the breakdown of burns is set out. Relation to earlier papers: Paper 506 showed that the same fuel yields more energy the faster the craft is moving when it burns——here the total $\Delta v$ of two burns straight to the target is compared with three burns going the long way, and the boundaries where the long way becomes cheaper are counted. The bi-elliptic transfer burns more than the Hohmann at the start, where the speed is greatest, and a little at the far apoapsis, where the speed is small. What is added is finding the boundaries $11.938765$ and $15.581719$ and the $3.304167$ beyond which the Hohmann transfer costs more than escape, finding the $B$ at which the bi-elliptic transfer becomes cheaper for $R=13$, checking the formula’s $\Delta v$ by numerical integration of the orbits, and placing the separator on the ratio of radii $R$.

1. Notation and the Question

Units: the radius of the starting circular orbit is $1$, and the speed on it is $1$. The radius of the target circular orbit is $R$, and the intermediate apoapsis of the bi-elliptic transfer is $B$.

The Hohmann transfer: it rides an ellipse with periapsis $1$ and apoapsis $R$. The total change of velocity is

$$H(R)=\left(\sqrt{\frac{2R}{1+R}}-1\right)+\left(\sqrt{\frac{1}{R}}-\sqrt{\frac{2}{R(1+R)}}\right).$$

The bi-elliptic transfer: it rides an ellipse with periapsis $1$ and apoapsis $B$, raises the periapsis to $R$ at $B$, and circularises at $R$. As $B$ is taken ever farther, the total approaches

$$H_\infty(R)=(\sqrt{2}-1)\left(1+\frac{1}{\sqrt{R}}\right).$$

Question: from which $R$ does the long way become cheaper?

2. Two Boundaries

Computation 2.1: $H(R)=H_\infty(R)$ at $R=\mathbf{11.938765}$. $H(R)$ reaches its maximum of $0.536258$ at $R=15.581719$. $H(R)$ begins to exceed the $\Delta v$ to escape to infinity from the start, $\sqrt{2}-1=0.414214$, at $R=3.304167$.

Observation 2.2 this is the core: the Hohmann $\Delta v$ does not keep rising as the target gets farther. Beyond $R=3.304167$, entering the far circular orbit needs more $\Delta v$ than escaping to infinity outright. Past $R=15.581719$, the Hohmann $\Delta v$ begins to fall, approaching $\sqrt{2}-1$ as $R$ grows without bound——this is why a path going out almost far enough to escape, then coming back with a small burn, can be cheaper than going straight there.

3. Between the Two Boundaries

Computation 3.1: for $400$ values of $R$ from just beyond $15.581719$ to $200$, moving $B$ through $120$ values from just outside $R$ out to $10^8$, the bi-elliptic transfer was cheaper in every case. At $R=13$, between the boundaries, the bi-elliptic transfer was cheaper only for $B$ beyond $48.904843$, and more expensive for nearer $B$.

Observation 3.2 between $11.94$ and $15.58$, how far out one goes decides which is cheaper. As $B$ approaches $R$, the bi-elliptic transfer turns into the Hohmann transfer (the third burn becomes $0$)——the gain of the long way appears only when one goes far enough.

4. Breakdown and Time

Computation 4.1: for $R=20$ and $B=100$, the $\Delta v$ of each burn was set out.

transferfirstsecondthirdtotal
Hohmann$0.380131$$0.154600$—$0.534731$
bi-elliptic ($B=100$)$0.407195$$0.043663$$0.065068$$0.515927$
bi-elliptic ($B\to\infty$)———$0.506835$

The bi-elliptic transfer is $0.018805$ cheaper. With the period of the starting orbit as $1$, the Hohmann transfer takes $17.011944$ and the bi-elliptic $411.813965$, $24.207343$ times as long.

Observation 4.2 the bi-elliptic transfer burns more at the start and a little far out. Its first burn, $0.407195$, is larger than the Hohmann $0.380131$, while the second, at apoapsis $100$, is only $0.043663$——it burns most on the starting circle, where the speed is greatest, and changes the shape of the orbit with little at the far apoapsis, where the speed is small.

5. Another Route: Numerical Integration of the Orbits

Computation 5.1: the orbits were integrated numerically, with burns of the formula’s $\Delta v$. The Hohmann transfer reached apoapsis $20.000000$, and after the second burn the radius stayed $20.000000$ through one revolution. The bi-elliptic transfer reached apoapsis $100.000000$, periapsis $20.000000$ after the second burn, and a radius of $20.000000$ through one revolution after the third. Both arrived within $10^{-5}$.

6. Real Sizes

Computation 6.1: with the Earth’s gravitational parameter, the start was a circular orbit at $300$ km altitude (radius $6678.137$ km).

target$R$Hohmannbi-elliptic ($B\to\infty$)
geostationary $42164$ km$6.313737$$3.892554$ km/s$4.473683$ km/s
Moon’s distance $384400$ km$57.560963$$3.936544$ km/s$3.621910$ km/s

Observation 6.2 to geostationary orbit ($R$ below $11.94$) the Hohmann transfer is cheaper; to the Moon’s distance ($R$ beyond $15.58$) the bi-elliptic transfer going far out is cheaper. The Hohmann $3.936544$ km/s to enter a circular orbit at the Moon’s distance is more than the $3.892554$ km/s to geostationary orbit——a far more distant target adds only a little $\Delta v$, and there the long way wins.

7. The Separator

Observation 7.1 the separator is the ratio of radii $R$. Below $11.94$ the Hohmann transfer wins, beyond $15.58$ the bi-elliptic transfer wins for every $B$, and in between it depends on how far out $B$ is——even between the same two circular orbits, “going straight is cheaper” belongs to a range of $R$.

Observation 7.2 beside Paper 506. Paper 506 showed that the same fuel yields more energy the faster the craft is moving when it burns——here the bi-elliptic transfer burns more than the Hohmann at the start, where the speed is greatest, and a little at the far apoapsis, where the speed is small. Where one burns shapes the total $\Delta v$.

Observation 7.3 what is not said. Real launch planning, time constraints and the gravity of other bodies are not treated, and the gain of the bi-elliptic transfer is not pinned to a single reason——what is shown is the values of the boundaries, the breakdown, and the check by numerical integration.

8. Roll Call and Closing

  1. equal at $R=\mathbf{11.938765}$ (Computation 2.1): with $B$ ever farther.
  2. Hohmann largest at $R=15.581719$ (Computation 2.1): $0.536258$.
  3. dearer than escape from $R=3.304167$ (Computation 2.1): exceeding $\sqrt{2}-1$.
  4. the Hohmann $\Delta v$ does not keep rising (Observation 2.2): the core.
  5. beyond $15.58$, bi-elliptic for every $B$ (Computation 3.1): $400$ by $120$ cases.
  6. at $R=13$, only for $B>48.904843$ (Computation 3.1): between the boundaries.
  7. $0.018805$ cheaper at $R=20$ (Computation 4.1): but $24.207343$ times as long.
  8. more at the start, a little far out (Observation 4.2): the breakdown.
  9. numerical integration arrives (Computation 5.1): within $10^{-5}$.
  10. where one burns matters (Observation 7.2): Paper 506.

Closing: Moving from a circular orbit of radius $1$ to one of radius $R$, once $R$ exceeds $11.938765$ the bi-elliptic transfer, going farther out first and coming back, needs a smaller total change of velocity than the direct Hohmann transfer, and beyond $15.581719$ it is cheaper however far out it goes. The Hohmann $\Delta v$ does not keep rising with distance: beyond $R=3.304167$ it costs more than escaping to infinity, and it peaks at $15.581719$——so a path going out almost far enough to escape and coming back can be cheaper. That the formula’s $\Delta v$ really arrives was checked by numerical integration of the orbits. The separator is the ratio of radii $R$. Placed among the earlier papers——the “where one burns” of Paper 506 shows up here in the breakdown of burns. To be honest——time constraints and real launch planning are not entered, and the gain of the bi-elliptic transfer is not pinned to a single reason.

References

  1. Hohmann, W. (1925). Die Erreichbarkeit der Himmelskörper. Oldenbourg.
  2. Yamagishi, Y. (2026). The Same Fuel Buys $3.90$ Times the Energy Depending on Where It Is Burned [Paper 506]. Zenodo. 10.5281/zenodo.22861840

Reference

Yamagishi, Y. (2026). For a Circular Orbit More Than 11.94 Times Farther Out, Going Even Farther First and Coming Back Needs Less Fuel [Paper 637]. YOI: https://www.yamagishiyuuki.com/yoi/637
DOI: 10.5281/zenodo.23014596
ORCID: 0009-0006-8587-3450

PDF (English) · PDF (Japanese) · Date proof of the English PDF (OpenTimestamps .ots) · SHA256SUMS

The record page for this paper (in Japanese, with the PDFs in each language and the date proof): YOI 637. This paper is a preprint and has not been peer reviewed.