A Question Worth Re-examining by Precision Experiment

Keying Guan and ChatGPT

Recently, while studying the relations among electromagnetic radiation, thermal motion, and macroscopic mechanical effects, we encountered an unpublished proposal developed about ten years ago by Dapeng Qian^[1] . The proposal raises a simple but unusual physical question:

Can microscopic thermal motion inside a solid produce a macroscopic inertial effect that has normally been neglected, causing the measured weight of the solid to decrease as its temperature increases?

We do not presently know whether Qian's proposed mechanism is correct.

However, after examining both his theoretical argument and several earlier experiments that motivated his analysis, we believe that the underlying experimental question deserves careful, independent, and substantially more precise investigation.

1. From Temperature to Microscopic Velocity

In classical statistical mechanics, temperature is related to microscopic motion. For a particle of mass in thermal equilibrium,

If we consider only the two components perpendicular to the local direction of gravity and denote their combined mean-square velocity by , then

or

These statistical-mechanical relations themselves are not new.

Historically, relations of this kind are closely connected with the kinetic theory of gases. We normally measure the temperature of a gas and infer the statistical scale of molecular motion; we do not measure the instantaneous velocities of all the molecules first and then determine the temperature.

For a solid, however, the microscopic physical situation is much more complicated. Atoms are not freely moving gas molecules. Their motions are constrained and correlated by the condensed-matter structure.

This makes the next step in Qian's proposal particularly unusual.

2. Qian's Proposed Step

Qian asks whether the microscopic transverse thermal motion represented statistically by  might have a normally neglected macroscopic inertial consequence.

For a body near the Earth's surface, he proposes an effective measured weight of the form

where  is the total mass of the body,  is the Earth's radius, and represents a characteristic particle mass associated with the material.

For a pure elemental solid, this may be viewed more transparently through the microscopic relation

The essential physical question can therefore be written schematically as

If Qian's proposed connection is valid, it leads immediately to a clear experimental prediction:

Moreover, the magnitude of the effect should depend on the material through its microscopic mass scale.

This is not an established physical relation. It is precisely the step that requires experimental testing.

3. The Proposal Did Not Arise in an Experimental Vacuum

An important point deserves emphasis.

Qian's proposal was developed in connection with previously reported experiments on the temperature dependence of measured weight.

Among them is the 2010 work of Liangzao Fan, Jinsong Feng, and Wuqing Liu, who reported measurements on several metals and found that their measured weights decreased as their temperatures increased^[2].

Qian went beyond noting the direction of the reported effect. In his manuscript he compared numerical predictions of his formula with previously reported experimental values.

For example, using one set of reported experimental conditions for copper, his calculation gives approximately

which he compared with the experimentally reported decrease of more than .

He also compared calculated and measured temperature coefficients for several metallic samples.

Thus, the historical logic is better represented as

This distinction is important. The proposal is not simply an abstract prediction waiting for its first experimental contact. Rather, it is one possible attempt to explain an already reported experimental phenomenon.

But that does not mean that the proposed mechanism has been established.

4. Observation Must Be Separated from Interpretation

The reported experiments have also received other interpretations.

For example, C. Y. Lo discussed the experiments of Fan, Feng, and Liu in a 2012 paper^[3]. Lo explicitly noted their reported finding that the measured weights of six metals decreased as temperature increased, but interpreted the phenomenon in terms of his proposed charge–mass interaction and an associated repulsive force.

Qian's interpretation is fundamentally different. He asks whether microscopic thermal motion itself can generate a previously neglected macroscopic inertial contribution.

We therefore think three logically distinct things must not be conflated:

Indeed, Lo himself concluded that the experiments deserved further testing and that others should check the results through the same or similar experiments.

This is also our position.

Before deciding which interpretation, if any, is correct, the experimental phenomenon itself should first be established with substantially greater precision.

5. Why Should the Experiment Be Repeated?

The central experimental question is remarkably simple:

If the answer is no, Qian's proposed mechanism would face direct experimental difficulty.

If the answer is yes, the next task would be much harder: one would have to determine whether the measured change can be explained by conventional effects such as air convection, buoyancy, thermal radiation, oxidation or other chemical changes, thermal deformation of the apparatus, temperature drift of the balance, or other instrumental and environmental effects.

Only after these possibilities have been quantitatively controlled would it be justified to discuss a new physical mechanism.

This is why a modern replication should preferably include a good vacuum environment or carefully controlled atmosphere, a high-resolution and highly stable balance, calibrated non-contact or rapid temperature measurement, multiple materials, control samples, and repeated heating–cooling cycles.

The crucial evidence would not be one isolated reading, but a reproducible relation

that survives reversal and independent replication.

In particular, we would look for

6. Our Own Preliminary Attempt

To gain some direct feeling for the problem, we recently performed a very simple exploratory experiment ourselves.

We first used a high-purity copper strip, rolled into a compact cylindrical sample. The room-temperature mass was approximately , and the sample was heated over roughly the range from  to .

The result was unsatisfactory. The heated copper surface visibly changed, suggesting oxidation or other surface modification, and the measured mass did not show the stable decrease we were looking for.

We then repeated the attempt using a high-purity aluminum sample of approximately . When its temperature was increased from about  to approximately , the electronic balance changed from  to .

The direction is consistent with .

However, the smallest displayed unit of the balance was itself . Therefore,

As a simple control, we also heated a nominal  calibration mass over a similar temperature interval; its displayed value remained .

These observations neither verify nor falsify Qian's proposal. Their main value was to show us how difficult the apparently simple experiment actually is.

Our experiment should therefore be regarded only as an exploratory attempt—not as evidence for a new physical effect.

7. What Kind of Experiment Is Really Needed?

Heating a macroscopic object changes many things simultaneously.

It can generate convection and buoyancy changes in the surrounding air. It can produce thermal radiation and temperature gradients in the balance. The specimen and its support expand. Surfaces can oxidize or adsorb/desorb material. Electronic balances themselves may respond to changing thermal conditions.

Consequently, the experiment ultimately required is not merely

It is closer to

Ideally, different laboratories should perform the experiment independently, using different balances, different heating methods, and several materials.

If a reproducible temperature-dependent signal remains substantially larger than all identified systematic effects, the phenomenon itself will deserve serious theoretical attention—irrespective of which present interpretation eventually proves correct.

8. Why Raise the Question Now?

The earlier experiments have existed in the literature for years, while Qian's proposal has remained unpublished for about a decade.

We therefore believe that the question should no longer remain confined to a small private discussion.

The purpose of this note is not to announce that Qian's theory has been proved.

Nor is it to ask readers to accept an unconventional mechanism.

It is to draw attention to a surprisingly elementary experimental question that appears capable, at least in principle, of a clear answer:

If careful modern experiments find no such effect beyond known uncertainties and systematic errors, that result is scientifically valuable.

If they find a reproducible effect but it can be fully explained by conventional thermal and instrumental physics, that too is scientifically valuable.

But if, after these effects are quantitatively excluded, a reproducible material-dependent anomaly remains, then the problem would deserve considerably deeper investigation.

The experiment should therefore come before commitment to any interpretation.

We invite interested experimentalists—especially those with access to precision balances, vacuum systems, and accurate temperature measurement—to repeat this deceptively simple experiment independently.

Nature, rather than any preferred interpretation, should decide.

References

[1] Dapeng Qian, A Previously Overlooked Macroscopic Inertial MechanismArising fromMicroscopic Thermal Motion and Its Measurable Consequences, to appear.

[2] Fan Liangzao , Feng Jinsong , Liu Wuqing, An experiment discovery about gravitational force changes in materials due to temperature variation, Engineering Sciences, 2010, 12(2): 9-11.

[3] C. Y. Lo, On the Weight Reduction of Metals Due to Temperature Increments, Global Journal of Science Frontier Research Physics and Space Sciences, Volume 12 Issue 7 Version 1.0 Year 2012

支持单位：