# Building a Computational Model of Planetary Motion (NKTg) and Validating It with NASA Mercury Data

**URL:** <https://forum.devtalk.com/t/building-a-computational-model-of-planetary-motion-nktg-and-validating-it-with-nasa-mercury-data/236200>\
**Category:** General Dev Chat\
**Created:** [3 March 2026 14:06 UTC](https://forum.devtalk.com/t/building-a-computational-model-of-planetary-motion-nktg-and-validating-it-with-nasa-mercury-data/236200 "2026-03-03T14:06:55Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![NKTgLaw](https://forum.devtalk.com/user_avatar/forum.devtalk.com/nktglaw/32/7590_2.png) [@NKTgLaw](https://forum.devtalk.com/u/NKTgLaw)\
**Post date:** [3 March 2026 14:06 UTC](https://forum.devtalk.com/t/building-a-computational-model-of-planetary-motion-nktg-and-validating-it-with-nasa-mercury-data/236200/1 "2026-03-03T14:06:55Z")

</div>

## 1. Problem Statement

Can planetary motion be simulated using only fundamental quantities such as:

- Position (x)

- Velocity (v)

- Mass (m)

without explicitly using force equations?

This post presents a computational implementation of the **NKTg model** , where orbital velocity is derived from a target constant:

> **NKTg₁ = x × p**  
> where p = m × v

The model is validated using publicly available data from NASA JPL Horizons for Mercury.

* * *

## 2. Reference Dataset (NASA 2024)

Actual orbital data of Mercury in 2024:

| Date | x (m) | v (m/s) | m (kg) | p = m·v |
| --- | --- | --- | --- | --- |
| 1/1/2024 | 5.16E+10 | 5.33E+04 | 3.30E+23 | 1.76E+28 |
| 4/1/2024 | 6.97E+10 | 3.90E+04 | 3.30E+23 | 1.29E+28 |
| 7/1/2024 | 5.36E+10 | 5.20E+04 | 3.30E+23 | 1.72E+28 |
| 10/1/2024 | 6.95E+10 | 3.92E+04 | 3.30E+23 | 1.30E+28 |
| 12/31/2024 | 4.64E+10 | 5.81E+04 | 3.30E+23 | 1.92E+28 |

From this dataset, the target constant is established:

> **NKTg₁ = 8.90E+38 NKTm**

* * *

## 3. Model Definition

Given:

- x (orbital position)

- m (mass)

- Target constant NKTg₁

We compute:

p=NKTg1xp = \frac{NKTg₁}{x}p=xNKTg1​​v=pmv = \frac{p}{m}v=mp​

No force expressions are used in the computation.

The model also integrates mass variation:

> dm/dt = -0.5 kg/s

This produces:

NKTg2=(dm/dt)×pNKTg₂ = (dm/dt) × pNKTg2​=(dm/dt)×p

which remains negative throughout the orbit.

* * *

## 4. 2025 Simulation (NKTg Model)

| Date | x (m) | v (m/s) | m (kg) | p = m·v |
| --- | --- | --- | --- | --- |
| 1/1/2025 | 5.16E+10 | 5.22E+04 | 3.301E+23 | 1.72E+28 |
| 4/1/2025 | 6.97E+10 | 3.87E+04 | 3.301E+23 | 1.28E+28 |
| 7/1/2025 | 5.49E+10 | 4.91E+04 | 3.301E+23 | 1.62E+28 |
| 10/1/2025 | 6.83E+10 | 3.95E+04 | 3.301E+23 | 1.30E+28 |
| 12/31/2025 | 4.61E+10 | 5.85E+04 | 3.301E+23 | 1.93E+28 |

* * *

## 5. Actual NASA Data (2025)

| Date | x (m) | v (m/s) | m (kg) |
| --- | --- | --- | --- |
| 1/1/2025 | 5.16E+10 | 5.34E+04 | 3.30E+23 |
| 4/1/2025 | 6.97E+10 | 3.89E+04 | 3.30E+23 |
| 7/1/2025 | 5.49E+10 | 5.04E+04 | 3.30E+23 |
| 1/10/2025 | 6.83E+10 | 3.98E+04 | 3.30E+23 |
| 12/31/2025 | 4.61E+10 | 5.89E+04 | 3.30E+23 |

* * *

## 6. Numerical Comparison

| Date | v – NKTg | v – NASA | Relative Error (%) |
| --- | --- | --- | --- |
| 1/1/2025 | 5.22E+04 | 53400 | -2.15461 |
| 4/1/2025 | 3.87E+04 | 38900 | -0.562786 |
| 7/1/2025 | 4.91E+04 | 50400 | -2.561981 |
| 1/10/2025 | 3.95E+04 | 39800 | -0.819213 |
| 12/31/2025 | 5.85E+04 | 58900 | -0.707806 |

**Average relative error: 1.3%**

* * *

## 7. Observations from a Computational Perspective

- The system is fully defined by interaction between x, v, and m.

- No force expressions are required in the computation process.

- The model is deterministic and reproducible across platforms.

- Error remains within ~1.3% compared to NASA JPL data.

* * *

## 8. Discussion

From a software engineering perspective, this approach represents:

- A constant-driven orbital simulation model.

- A reduced-parameter dynamical system.

- A reproducible computational experiment based on publicly available data.

The implementation can be reproduced in any language supporting floating-point arithmetic.

* * *

If anyone is interested, I can share a minimal reproducible implementation (Python / C++ / Go) for benchmarking and independent validation.
